diff --git "a/checkpoint-22932/trainer_state.json" "b/checkpoint-22932/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-22932/trainer_state.json"
@@ -0,0 +1,160545 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.7360708288810869,
+  "eval_steps": 500,
+  "global_step": 22932,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.1258,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.8013,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 6e-06,
+      "loss": 1.807,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.2332,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1e-05,
+      "loss": 2.1132,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.2e-05,
+      "loss": 1.915,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.4e-05,
+      "loss": 2.0501,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.3626,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.8e-05,
+      "loss": 1.9588,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2e-05,
+      "loss": 1.8198,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.9999999982026183e-05,
+      "loss": 1.9872,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.9999999928104733e-05,
+      "loss": 2.1587,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.9999999838235645e-05,
+      "loss": 1.6889,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.9999999712418923e-05,
+      "loss": 2.1361,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.9999999550654566e-05,
+      "loss": 2.0777,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.9999999352942575e-05,
+      "loss": 1.757,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.9999999119282956e-05,
+      "loss": 2.0809,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.9999998849675703e-05,
+      "loss": 1.879,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999998544120815e-05,
+      "loss": 2.2581,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.99999982026183e-05,
+      "loss": 1.7315,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.9999997825168157e-05,
+      "loss": 1.9142,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.999999741177039e-05,
+      "loss": 2.0022,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.9999996962424992e-05,
+      "loss": 2.3471,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9999996477131973e-05,
+      "loss": 2.0283,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.9999995955891326e-05,
+      "loss": 1.706,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9999995398703065e-05,
+      "loss": 2.1073,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.9999994805567183e-05,
+      "loss": 1.6502,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9999994176483684e-05,
+      "loss": 1.995,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.9999993511452574e-05,
+      "loss": 2.149,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.9999992810473846e-05,
+      "loss": 1.9025,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9999992073547515e-05,
+      "loss": 1.8489,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.9999991300673573e-05,
+      "loss": 1.8234,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.999999049185203e-05,
+      "loss": 2.6627,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.9999989647082887e-05,
+      "loss": 2.2688,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.9999988766366143e-05,
+      "loss": 2.1593,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.999998784970181e-05,
+      "loss": 2.1343,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.9999986897089877e-05,
+      "loss": 1.7185,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.9999985908530362e-05,
+      "loss": 1.9871,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.999998488402326e-05,
+      "loss": 2.0824,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.9999983823568578e-05,
+      "loss": 1.7671,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.999998272716632e-05,
+      "loss": 1.8736,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.9999981594816486e-05,
+      "loss": 1.8292,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.9999980426519087e-05,
+      "loss": 1.8383,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.999997922227412e-05,
+      "loss": 1.7324,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.99999779820816e-05,
+      "loss": 1.7436,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9999976705941517e-05,
+      "loss": 2.2156,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.999997539385388e-05,
+      "loss": 1.7598,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.9999974045818703e-05,
+      "loss": 2.0896,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.999997266183598e-05,
+      "loss": 1.6664,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9999971241905724e-05,
+      "loss": 2.9754,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9999969786027934e-05,
+      "loss": 1.9749,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9999968294202617e-05,
+      "loss": 2.2035,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.9999966766429782e-05,
+      "loss": 2.019,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9999965202709427e-05,
+      "loss": 2.0146,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999963603041563e-05,
+      "loss": 1.9535,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.9999961967426197e-05,
+      "loss": 1.85,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9999960295863332e-05,
+      "loss": 1.8697,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.9999958588352975e-05,
+      "loss": 2.2409,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.999995684489513e-05,
+      "loss": 1.7714,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999955065489808e-05,
+      "loss": 1.8298,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999953250137008e-05,
+      "loss": 1.811,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9999951398836743e-05,
+      "loss": 2.3906,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.999994951158902e-05,
+      "loss": 1.7484,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.999994758839384e-05,
+      "loss": 2.1229,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9999945629251216e-05,
+      "loss": 2.1949,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9999943634161152e-05,
+      "loss": 1.8442,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9999941603123654e-05,
+      "loss": 1.7307,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.999993953613873e-05,
+      "loss": 1.8975,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9999937433206394e-05,
+      "loss": 2.0353,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.9999935294326643e-05,
+      "loss": 1.9755,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.999993311949949e-05,
+      "loss": 1.9375,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.9999930908724942e-05,
+      "loss": 2.1417,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.9999928662003006e-05,
+      "loss": 1.8652,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9999926379333696e-05,
+      "loss": 1.9959,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9999924060717013e-05,
+      "loss": 2.3449,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.999992170615297e-05,
+      "loss": 2.1993,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.999991931564157e-05,
+      "loss": 1.8305,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9999916889182826e-05,
+      "loss": 2.0184,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.9999914426776744e-05,
+      "loss": 2.0368,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.999991192842334e-05,
+      "loss": 1.7324,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.999990939412261e-05,
+      "loss": 2.0312,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9999906823874576e-05,
+      "loss": 1.5432,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.9999904217679243e-05,
+      "loss": 2.0401,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9999901575536615e-05,
+      "loss": 2.0799,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.9999898897446708e-05,
+      "loss": 1.728,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.999989618340953e-05,
+      "loss": 2.0258,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.9999893433425088e-05,
+      "loss": 1.9614,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9999890647493395e-05,
+      "loss": 2.1852,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.999988782561446e-05,
+      "loss": 1.9545,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9999884967788293e-05,
+      "loss": 2.1906,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9999882074014903e-05,
+      "loss": 2.1311,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9999879144294306e-05,
+      "loss": 1.7983,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9999876178626505e-05,
+      "loss": 1.905,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9999873177011516e-05,
+      "loss": 1.968,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9999870139449348e-05,
+      "loss": 2.0054,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9999867065940012e-05,
+      "loss": 2.0835,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.9999863956483513e-05,
+      "loss": 1.791,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999860811079873e-05,
+      "loss": 1.5302,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9999857629729098e-05,
+      "loss": 1.8508,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.99998544124312e-05,
+      "loss": 2.1439,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.999985115918619e-05,
+      "loss": 2.0763,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.999984786999408e-05,
+      "loss": 1.657,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.999984454485488e-05,
+      "loss": 1.8415,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.9999841183768606e-05,
+      "loss": 2.0349,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.9999837786735266e-05,
+      "loss": 1.721,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.9999834353754878e-05,
+      "loss": 1.5898,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9999830884827445e-05,
+      "loss": 1.8272,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9999827379952988e-05,
+      "loss": 1.8106,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9999823839131516e-05,
+      "loss": 2.1147,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999982026236304e-05,
+      "loss": 2.3803,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9999816649647577e-05,
+      "loss": 1.8948,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.999981300098514e-05,
+      "loss": 2.1835,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.999980931637574e-05,
+      "loss": 2.0579,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9999805595819385e-05,
+      "loss": 1.7897,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9999801839316098e-05,
+      "loss": 2.2369,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9999798046865888e-05,
+      "loss": 2.2906,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.999979421846877e-05,
+      "loss": 1.7347,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.9999790354124753e-05,
+      "loss": 1.6501,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25,
+      "learning_rate": 1.999978645383386e-05,
+      "loss": 1.7091,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9999782517596093e-05,
+      "loss": 2.0927,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.999977854541148e-05,
+      "loss": 1.959,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9999774537280022e-05,
+      "loss": 1.81,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.999977049320174e-05,
+      "loss": 1.8844,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9999766413176652e-05,
+      "loss": 2.6295,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.9999762297204766e-05,
+      "loss": 1.8338,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25,
+      "learning_rate": 1.99997581452861e-05,
+      "loss": 1.7116,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9999753957420665e-05,
+      "loss": 2.3333,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9999749733608485e-05,
+      "loss": 1.9274,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9999745473849564e-05,
+      "loss": 1.888,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9999741178143927e-05,
+      "loss": 2.2129,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9999736846491588e-05,
+      "loss": 1.9963,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9999732478892552e-05,
+      "loss": 1.9378,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.9999728075346848e-05,
+      "loss": 2.1563,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9999723635854485e-05,
+      "loss": 1.5033,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25,
+      "learning_rate": 1.999971916041548e-05,
+      "loss": 1.9651,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999971464902985e-05,
+      "loss": 1.9047,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.999971010169761e-05,
+      "loss": 1.9141,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9999705518418778e-05,
+      "loss": 1.7182,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999970089919337e-05,
+      "loss": 1.9985,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.99996962440214e-05,
+      "loss": 2.0418,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.999969155290289e-05,
+      "loss": 2.0227,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.999968682583785e-05,
+      "loss": 2.0814,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9999682062826302e-05,
+      "loss": 1.7386,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.9999677263868264e-05,
+      "loss": 1.6724,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.999967242896375e-05,
+      "loss": 1.7238,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9999667558112778e-05,
+      "loss": 1.9934,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.9999662651315367e-05,
+      "loss": 1.6736,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999965770857153e-05,
+      "loss": 2.0692,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.999965272988129e-05,
+      "loss": 2.0752,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9999647715244666e-05,
+      "loss": 1.9654,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9999642664661674e-05,
+      "loss": 1.949,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9999637578132328e-05,
+      "loss": 2.1258,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999963245565665e-05,
+      "loss": 1.6346,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9999627297234657e-05,
+      "loss": 1.6057,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.9999622102866376e-05,
+      "loss": 1.7838,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.999961687255181e-05,
+      "loss": 2.042,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.999961160629099e-05,
+      "loss": 1.8385,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.999960630408393e-05,
+      "loss": 1.8294,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.999960096593065e-05,
+      "loss": 1.806,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.999959559183117e-05,
+      "loss": 1.8704,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999959018178551e-05,
+      "loss": 2.0,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9999584735793686e-05,
+      "loss": 2.0034,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.999957925385572e-05,
+      "loss": 1.8676,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9999573735971634e-05,
+      "loss": 2.1761,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9999568182141442e-05,
+      "loss": 1.9497,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.999956259236517e-05,
+      "loss": 1.9955,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9999556966642833e-05,
+      "loss": 1.9991,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9999551304974453e-05,
+      "loss": 2.1884,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9999545607360052e-05,
+      "loss": 1.7786,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9999539873799653e-05,
+      "loss": 1.8477,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9999534104293266e-05,
+      "loss": 1.9506,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9999528298840922e-05,
+      "loss": 1.8594,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.999952245744264e-05,
+      "loss": 2.1033,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.999951658009844e-05,
+      "loss": 2.1137,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999951066680834e-05,
+      "loss": 1.9811,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9999504717572365e-05,
+      "loss": 2.0154,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9999498732390533e-05,
+      "loss": 2.0408,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9999492711262875e-05,
+      "loss": 1.8284,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.99994866541894e-05,
+      "loss": 1.8456,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.999948056117013e-05,
+      "loss": 1.7695,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9999474432205103e-05,
+      "loss": 2.0752,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9999468267294324e-05,
+      "loss": 1.9335,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9999462066437822e-05,
+      "loss": 1.978,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.999945582963562e-05,
+      "loss": 1.6673,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25,
+      "learning_rate": 1.999944955688774e-05,
+      "loss": 1.9261,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.99994432481942e-05,
+      "loss": 1.8498,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.999943690355503e-05,
+      "loss": 1.8874,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.9999430522970246e-05,
+      "loss": 1.834,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9999424106439877e-05,
+      "loss": 1.7158,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999941765396394e-05,
+      "loss": 2.1064,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9999411165542466e-05,
+      "loss": 1.8814,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9999404641175472e-05,
+      "loss": 2.2142,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.999939808086298e-05,
+      "loss": 1.9295,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999939148460502e-05,
+      "loss": 1.8338,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.999938485240161e-05,
+      "loss": 2.092,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.999937818425278e-05,
+      "loss": 1.7572,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9999371480158544e-05,
+      "loss": 2.0562,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999936474011894e-05,
+      "loss": 1.904,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9999357964133978e-05,
+      "loss": 2.0415,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9999351152203692e-05,
+      "loss": 1.8919,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.99993443043281e-05,
+      "loss": 1.7855,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9999337420507238e-05,
+      "loss": 1.8338,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9999330500741117e-05,
+      "loss": 1.8677,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9999323545029767e-05,
+      "loss": 1.8132,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9999316553373215e-05,
+      "loss": 2.1013,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9999309525771485e-05,
+      "loss": 2.1769,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9999302462224597e-05,
+      "loss": 2.1145,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999929536273259e-05,
+      "loss": 1.9649,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9999288227295475e-05,
+      "loss": 1.9946,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.9999281055913283e-05,
+      "loss": 1.7902,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9999273848586046e-05,
+      "loss": 2.1407,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.999926660531378e-05,
+      "loss": 2.0247,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9999259326096515e-05,
+      "loss": 1.7965,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.999925201093428e-05,
+      "loss": 2.0347,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.99992446598271e-05,
+      "loss": 2.1978,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9999237272774995e-05,
+      "loss": 2.0722,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9999229849778e-05,
+      "loss": 2.4401,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.999922239083614e-05,
+      "loss": 2.2132,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999921489594944e-05,
+      "loss": 1.8941,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999920736511792e-05,
+      "loss": 1.8148,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.999919979834162e-05,
+      "loss": 1.9332,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9999192195620566e-05,
+      "loss": 1.7171,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9999184556954777e-05,
+      "loss": 2.2056,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.999917688234428e-05,
+      "loss": 1.7519,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.999916917178911e-05,
+      "loss": 1.9936,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9999161425289293e-05,
+      "loss": 1.7259,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.9999153642844855e-05,
+      "loss": 2.0016,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9999145824455822e-05,
+      "loss": 1.7148,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9999137970122228e-05,
+      "loss": 2.0479,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9999130079844095e-05,
+      "loss": 1.9385,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9999122153621453e-05,
+      "loss": 1.9254,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9999114191454337e-05,
+      "loss": 2.3394,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9999106193342764e-05,
+      "loss": 1.924,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9999098159286773e-05,
+      "loss": 1.8318,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9999090089286386e-05,
+      "loss": 1.6599,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9999081983341634e-05,
+      "loss": 2.2471,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.999907384145255e-05,
+      "loss": 2.0045,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999906566361916e-05,
+      "loss": 2.0395,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9999057449841494e-05,
+      "loss": 1.8034,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999904920011958e-05,
+      "loss": 1.8612,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9999040914453452e-05,
+      "loss": 2.1539,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.999903259284313e-05,
+      "loss": 1.7985,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9999024235288655e-05,
+      "loss": 2.0302,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.999901584179005e-05,
+      "loss": 2.1209,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9999007412347353e-05,
+      "loss": 1.9222,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9998998946960582e-05,
+      "loss": 2.1212,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.999899044562978e-05,
+      "loss": 1.9541,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.9998981908354972e-05,
+      "loss": 2.0,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9998973335136187e-05,
+      "loss": 2.1428,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9998964725973458e-05,
+      "loss": 1.782,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9998956080866817e-05,
+      "loss": 1.7536,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.999894739981629e-05,
+      "loss": 1.9021,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9998938682821912e-05,
+      "loss": 1.8634,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999892992988372e-05,
+      "loss": 2.1437,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.999892114100173e-05,
+      "loss": 2.1187,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25,
+      "learning_rate": 1.999891231617599e-05,
+      "loss": 1.9225,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9998903455406523e-05,
+      "loss": 1.7131,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9998894558693363e-05,
+      "loss": 1.7773,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.999888562603654e-05,
+      "loss": 2.0019,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.999887665743609e-05,
+      "loss": 1.9303,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9998867652892043e-05,
+      "loss": 1.7722,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999885861240443e-05,
+      "loss": 1.7296,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9998849535973287e-05,
+      "loss": 2.1225,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9998840423598644e-05,
+      "loss": 1.8992,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999883127528053e-05,
+      "loss": 1.7513,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.999882209101899e-05,
+      "loss": 2.0494,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9998812870814043e-05,
+      "loss": 2.181,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9998803614665735e-05,
+      "loss": 1.6803,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9998794322574086e-05,
+      "loss": 1.6499,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.999878499453914e-05,
+      "loss": 1.7333,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9998775630560922e-05,
+      "loss": 1.99,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9998766230639475e-05,
+      "loss": 1.5802,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9998756794774825e-05,
+      "loss": 1.7364,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.999874732296701e-05,
+      "loss": 1.72,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9998737815216062e-05,
+      "loss": 1.7472,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9998728271522017e-05,
+      "loss": 2.0934,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9998718691884912e-05,
+      "loss": 1.8514,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9998709076304773e-05,
+      "loss": 1.9322,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9998699424781642e-05,
+      "loss": 1.9245,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9998689737315552e-05,
+      "loss": 1.9287,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9998680013906533e-05,
+      "loss": 1.8705,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9998670254554627e-05,
+      "loss": 2.0223,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9998660459259867e-05,
+      "loss": 1.9271,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9998650628022288e-05,
+      "loss": 2.2253,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9998640760841922e-05,
+      "loss": 1.8108,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9998630857718807e-05,
+      "loss": 2.1567,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.999862091865298e-05,
+      "loss": 1.9705,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.999861094364447e-05,
+      "loss": 2.2423,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9998600932693327e-05,
+      "loss": 2.2131,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9998590885799573e-05,
+      "loss": 1.9679,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.999858080296325e-05,
+      "loss": 2.1468,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9998570684184395e-05,
+      "loss": 1.9925,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9998560529463042e-05,
+      "loss": 1.7714,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.999855033879923e-05,
+      "loss": 2.0329,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.999854011219299e-05,
+      "loss": 1.6006,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9998529849644364e-05,
+      "loss": 1.7952,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9998519551153392e-05,
+      "loss": 1.9144,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9998509216720105e-05,
+      "loss": 2.0384,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.999849884634454e-05,
+      "loss": 1.9255,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9998488440026737e-05,
+      "loss": 2.1372,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9998477997766735e-05,
+      "loss": 1.9239,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9998467519564567e-05,
+      "loss": 1.816,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9998457005420273e-05,
+      "loss": 1.7692,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.999844645533389e-05,
+      "loss": 1.9715,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9998435869305456e-05,
+      "loss": 2.1029,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9998425247335013e-05,
+      "loss": 1.8033,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.999841458942259e-05,
+      "loss": 2.2095,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9998403895568236e-05,
+      "loss": 1.8031,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9998393165771987e-05,
+      "loss": 2.1995,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9998382400033873e-05,
+      "loss": 2.074,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.999837159835394e-05,
+      "loss": 2.0706,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999836076073223e-05,
+      "loss": 1.7051,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9998349887168772e-05,
+      "loss": 2.0908,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9998338977663614e-05,
+      "loss": 2.1128,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.999832803221679e-05,
+      "loss": 2.0124,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9998317050828344e-05,
+      "loss": 2.0174,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9998306033498308e-05,
+      "loss": 1.6634,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.999829498022673e-05,
+      "loss": 1.8091,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9998283891013647e-05,
+      "loss": 1.9081,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9998272765859098e-05,
+      "loss": 1.9387,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.999826160476312e-05,
+      "loss": 1.8607,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9998250407725758e-05,
+      "loss": 1.841,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999823917474705e-05,
+      "loss": 1.892,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9998227905827034e-05,
+      "loss": 1.9921,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9998216600965758e-05,
+      "loss": 1.9404,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9998205260163254e-05,
+      "loss": 2.1486,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.999819388341957e-05,
+      "loss": 1.7653,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999818247073474e-05,
+      "loss": 2.0174,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.999817102210881e-05,
+      "loss": 1.9296,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999815953754182e-05,
+      "loss": 1.9077,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.999814801703381e-05,
+      "loss": 1.8855,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9998136460584823e-05,
+      "loss": 1.7558,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.99981248681949e-05,
+      "loss": 2.0463,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9998113239864084e-05,
+      "loss": 2.0383,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9998101575592413e-05,
+      "loss": 1.9619,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9998089875379932e-05,
+      "loss": 1.9511,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999807813922668e-05,
+      "loss": 1.9306,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9998066367132707e-05,
+      "loss": 2.0578,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9998054559098047e-05,
+      "loss": 1.6276,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.9998042715122747e-05,
+      "loss": 1.8071,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9998030835206845e-05,
+      "loss": 2.1289,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9998018919350387e-05,
+      "loss": 2.032,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9998006967553417e-05,
+      "loss": 2.0088,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9997994979815978e-05,
+      "loss": 1.8226,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9997982956138107e-05,
+      "loss": 1.9789,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9997970896519856e-05,
+      "loss": 1.9128,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9997958800961262e-05,
+      "loss": 2.0812,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.999794666946237e-05,
+      "loss": 1.9722,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9997934502023224e-05,
+      "loss": 1.6736,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9997922298643872e-05,
+      "loss": 2.0565,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.999791005932435e-05,
+      "loss": 1.7054,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9997897784064703e-05,
+      "loss": 2.0867,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9997885472864982e-05,
+      "loss": 1.711,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9997873125725226e-05,
+      "loss": 1.8679,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9997860742645482e-05,
+      "loss": 1.7759,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.999784832362579e-05,
+      "loss": 2.1024,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9997835868666202e-05,
+      "loss": 2.0507,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9997823377766753e-05,
+      "loss": 1.9385,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.99978108509275e-05,
+      "loss": 2.0545,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9997798288148477e-05,
+      "loss": 2.1714,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9997785689429734e-05,
+      "loss": 1.6993,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9997773054771315e-05,
+      "loss": 2.3004,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999776038417327e-05,
+      "loss": 1.8902,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.999774767763564e-05,
+      "loss": 1.5517,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9997734935158468e-05,
+      "loss": 1.8819,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9997722156741807e-05,
+      "loss": 1.6375,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9997709342385697e-05,
+      "loss": 2.1043,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9997696492090187e-05,
+      "loss": 1.8393,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9997683605855322e-05,
+      "loss": 1.7219,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9997670683681147e-05,
+      "loss": 2.0111,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9997657725567714e-05,
+      "loss": 2.0971,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9997644731515065e-05,
+      "loss": 1.9806,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9997631701523246e-05,
+      "loss": 1.9649,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9997618635592306e-05,
+      "loss": 1.8427,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9997605533722294e-05,
+      "loss": 1.9064,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.999759239591325e-05,
+      "loss": 2.1141,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9997579222165228e-05,
+      "loss": 2.0995,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9997566012478273e-05,
+      "loss": 2.0172,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9997552766852434e-05,
+      "loss": 1.751,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9997539485287754e-05,
+      "loss": 2.0661,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9997526167784285e-05,
+      "loss": 2.1808,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9997512814342076e-05,
+      "loss": 1.6732,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.999749942496117e-05,
+      "loss": 1.9122,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9997485999641618e-05,
+      "loss": 1.8665,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.999747253838347e-05,
+      "loss": 1.8985,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.999745904118677e-05,
+      "loss": 2.0437,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.999744550805157e-05,
+      "loss": 2.0224,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9997431938977917e-05,
+      "loss": 2.2041,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9997418333965862e-05,
+      "loss": 1.9332,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9997404693015453e-05,
+      "loss": 2.0454,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999739101612674e-05,
+      "loss": 1.9887,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9997377303299767e-05,
+      "loss": 2.0738,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.999736355453459e-05,
+      "loss": 2.2476,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9997349769831254e-05,
+      "loss": 2.0195,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9997335949189808e-05,
+      "loss": 1.8642,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9997322092610306e-05,
+      "loss": 1.9688,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.999730820009279e-05,
+      "loss": 1.9211,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9997294271637323e-05,
+      "loss": 1.9402,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9997280307243944e-05,
+      "loss": 1.7436,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9997266306912708e-05,
+      "loss": 1.7943,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9997252270643664e-05,
+      "loss": 2.0903,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9997238198436858e-05,
+      "loss": 1.8098,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9997224090292348e-05,
+      "loss": 2.079,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9997209946210182e-05,
+      "loss": 2.0743,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.999719576619041e-05,
+      "loss": 2.0086,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9997181550233087e-05,
+      "loss": 1.5996,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9997167298338257e-05,
+      "loss": 2.0652,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9997153010505974e-05,
+      "loss": 1.7714,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.999713868673629e-05,
+      "loss": 1.8018,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9997124327029258e-05,
+      "loss": 1.7168,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9997109931384927e-05,
+      "loss": 1.9086,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9997095499803353e-05,
+      "loss": 2.3703,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9997081032284583e-05,
+      "loss": 1.9006,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9997066528828672e-05,
+      "loss": 1.6944,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.999705198943567e-05,
+      "loss": 2.3348,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9997037414105632e-05,
+      "loss": 1.8672,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9997022802838605e-05,
+      "loss": 1.8059,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9997008155634647e-05,
+      "loss": 1.7887,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9996993472493812e-05,
+      "loss": 1.9179,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9996978753416147e-05,
+      "loss": 1.9034,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9996963998401706e-05,
+      "loss": 2.0982,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9996949207450548e-05,
+      "loss": 2.1444,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.999693438056272e-05,
+      "loss": 2.0551,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9996919517738274e-05,
+      "loss": 1.8969,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9996904618977267e-05,
+      "loss": 2.2586,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9996889684279756e-05,
+      "loss": 1.9002,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.9996874713645787e-05,
+      "loss": 1.7837,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.999685970707542e-05,
+      "loss": 1.9614,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9996844664568706e-05,
+      "loss": 2.21,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.99968295861257e-05,
+      "loss": 1.9528,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9996814471746455e-05,
+      "loss": 1.9331,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9996799321431028e-05,
+      "loss": 2.0927,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.999678413517947e-05,
+      "loss": 2.0182,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9996768912991834e-05,
+      "loss": 2.1925,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9996753654868185e-05,
+      "loss": 1.8878,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9996738360808566e-05,
+      "loss": 2.1166,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9996723030813037e-05,
+      "loss": 1.863,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9996707664881653e-05,
+      "loss": 1.9847,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.999669226301447e-05,
+      "loss": 2.0039,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9996676825211544e-05,
+      "loss": 1.7839,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9996661351472924e-05,
+      "loss": 2.1588,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9996645841798674e-05,
+      "loss": 1.731,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9996630296188848e-05,
+      "loss": 2.1226,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.99966147146435e-05,
+      "loss": 1.6108,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.999659909716268e-05,
+      "loss": 1.9088,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9996583443746457e-05,
+      "loss": 2.2507,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.999656775439488e-05,
+      "loss": 1.8619,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9996552029108006e-05,
+      "loss": 2.2089,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9996536267885885e-05,
+      "loss": 1.7759,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9996520470728588e-05,
+      "loss": 1.9723,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9996504637636162e-05,
+      "loss": 1.9756,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9996488768608665e-05,
+      "loss": 2.2497,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9996472863646154e-05,
+      "loss": 1.9455,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9996456922748687e-05,
+      "loss": 2.1995,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9996440945916326e-05,
+      "loss": 1.8293,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.999642493314912e-05,
+      "loss": 2.2772,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.999640888444713e-05,
+      "loss": 1.5146,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999639279981042e-05,
+      "loss": 1.6934,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9996376679239032e-05,
+      "loss": 1.7476,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9996360522733042e-05,
+      "loss": 1.8648,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9996344330292496e-05,
+      "loss": 2.0939,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.999632810191746e-05,
+      "loss": 1.982,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9996311837607985e-05,
+      "loss": 1.9569,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9996295537364136e-05,
+      "loss": 2.2778,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9996279201185964e-05,
+      "loss": 1.7772,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9996262829073534e-05,
+      "loss": 2.208,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9996246421026908e-05,
+      "loss": 2.0283,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9996229977046134e-05,
+      "loss": 1.5991,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999621349713128e-05,
+      "loss": 1.8888,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.99961969812824e-05,
+      "loss": 2.0867,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9996180429499557e-05,
+      "loss": 1.8767,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9996163841782812e-05,
+      "loss": 2.1289,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.999614721813222e-05,
+      "loss": 1.6679,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9996130558547842e-05,
+      "loss": 1.8021,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.999611386302974e-05,
+      "loss": 2.0548,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9996097131577974e-05,
+      "loss": 1.747,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9996080364192602e-05,
+      "loss": 2.009,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9996063560873683e-05,
+      "loss": 1.9647,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.9996046721621282e-05,
+      "loss": 1.8054,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9996029846435453e-05,
+      "loss": 2.0562,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9996012935316263e-05,
+      "loss": 1.8467,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9995995988263774e-05,
+      "loss": 2.2326,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.999597900527804e-05,
+      "loss": 1.8744,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9995961986359126e-05,
+      "loss": 2.1177,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9995944931507096e-05,
+      "loss": 2.0258,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9995927840722e-05,
+      "loss": 2.0904,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9995910714003916e-05,
+      "loss": 1.9808,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9995893551352892e-05,
+      "loss": 1.8071,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9995876352768996e-05,
+      "loss": 2.3563,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9995859118252285e-05,
+      "loss": 1.9476,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9995841847802827e-05,
+      "loss": 1.6587,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9995824541420683e-05,
+      "loss": 1.8243,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999580719910591e-05,
+      "loss": 1.8667,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.999578982085858e-05,
+      "loss": 1.8325,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9995772406678745e-05,
+      "loss": 1.5781,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9995754956566474e-05,
+      "loss": 2.0856,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999573747052183e-05,
+      "loss": 1.8656,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.999571994854487e-05,
+      "loss": 1.7105,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9995702390635662e-05,
+      "loss": 2.2442,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9995684796794266e-05,
+      "loss": 2.2047,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.999566716702075e-05,
+      "loss": 1.9171,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9995649501315172e-05,
+      "loss": 2.1853,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.99956317996776e-05,
+      "loss": 1.8549,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.999561406210809e-05,
+      "loss": 2.1751,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.999559628860672e-05,
+      "loss": 2.2348,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.999557847917354e-05,
+      "loss": 1.935,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.999556063380862e-05,
+      "loss": 1.7636,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9995542752512022e-05,
+      "loss": 1.8924,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9995524835283815e-05,
+      "loss": 2.0049,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9995506882124057e-05,
+      "loss": 2.0632,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.999548889303282e-05,
+      "loss": 2.0137,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9995470868010162e-05,
+      "loss": 1.9184,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9995452807056147e-05,
+      "loss": 1.73,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9995434710170846e-05,
+      "loss": 2.0416,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9995416577354322e-05,
+      "loss": 1.8906,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9995398408606637e-05,
+      "loss": 1.7802,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.999538020392786e-05,
+      "loss": 2.0497,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9995361963318053e-05,
+      "loss": 2.2303,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9995343686777287e-05,
+      "loss": 1.8621,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.999532537430562e-05,
+      "loss": 1.9213,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9995307025903125e-05,
+      "loss": 2.0571,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9995288641569865e-05,
+      "loss": 2.6279,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9995270221305903e-05,
+      "loss": 1.8742,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9995251765111313e-05,
+      "loss": 2.0281,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9995233272986154e-05,
+      "loss": 1.8504,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9995214744930498e-05,
+      "loss": 2.033,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9995196180944405e-05,
+      "loss": 1.7853,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.999517758102795e-05,
+      "loss": 1.8236,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.999515894518119e-05,
+      "loss": 1.7435,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.99951402734042e-05,
+      "loss": 1.9173,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9995121565697048e-05,
+      "loss": 1.9992,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.999510282205979e-05,
+      "loss": 1.9527,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9995084042492506e-05,
+      "loss": 1.7687,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9995065226995257e-05,
+      "loss": 1.8125,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.9995046375568114e-05,
+      "loss": 1.5668,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9995027488211144e-05,
+      "loss": 1.9381,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9995008564924412e-05,
+      "loss": 1.7568,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.999498960570799e-05,
+      "loss": 1.8043,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.999497061056194e-05,
+      "loss": 1.459,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9994951579486336e-05,
+      "loss": 1.5666,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9994932512481247e-05,
+      "loss": 1.9636,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9994913409546738e-05,
+      "loss": 1.9235,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9994894270682877e-05,
+      "loss": 1.9875,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999487509588974e-05,
+      "loss": 1.8323,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9994855885167384e-05,
+      "loss": 2.056,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.999483663851589e-05,
+      "loss": 1.6506,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.999481735593532e-05,
+      "loss": 1.7102,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9994798037425746e-05,
+      "loss": 1.5048,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9994778682987236e-05,
+      "loss": 2.0331,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.999475929261986e-05,
+      "loss": 2.0204,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.999473986632369e-05,
+      "loss": 2.0589,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9994720404098794e-05,
+      "loss": 2.1543,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9994700905945237e-05,
+      "loss": 1.9555,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9994681371863098e-05,
+      "loss": 2.0459,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.9994661801852444e-05,
+      "loss": 1.9843,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9994642195913342e-05,
+      "loss": 1.9211,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9994622554045867e-05,
+      "loss": 1.9968,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9994602876250087e-05,
+      "loss": 1.9759,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.999458316252607e-05,
+      "loss": 2.1961,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9994563412873896e-05,
+      "loss": 1.785,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9994543627293628e-05,
+      "loss": 1.8724,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.999452380578534e-05,
+      "loss": 1.9935,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9994503948349097e-05,
+      "loss": 1.8172,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.999448405498498e-05,
+      "loss": 2.1128,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9994464125693057e-05,
+      "loss": 2.1195,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9994444160473397e-05,
+      "loss": 2.28,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.999442415932608e-05,
+      "loss": 2.109,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9994404122251165e-05,
+      "loss": 2.0535,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9994384049248734e-05,
+      "loss": 2.0452,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9994363940318852e-05,
+      "loss": 2.0251,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.99943437954616e-05,
+      "loss": 1.8555,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9994323614677043e-05,
+      "loss": 2.0936,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9994303397965258e-05,
+      "loss": 2.2546,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9994283145326313e-05,
+      "loss": 2.7214,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9994262856760286e-05,
+      "loss": 1.718,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.999424253226725e-05,
+      "loss": 1.7934,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.999422217184727e-05,
+      "loss": 1.8819,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.999420177550043e-05,
+      "loss": 2.0089,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9994181343226794e-05,
+      "loss": 1.6305,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9994160875026444e-05,
+      "loss": 1.8453,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9994140370899443e-05,
+      "loss": 1.9405,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9994119830845874e-05,
+      "loss": 2.038,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.999409925486581e-05,
+      "loss": 1.9131,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.999407864295932e-05,
+      "loss": 1.6785,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9994057995126485e-05,
+      "loss": 1.9999,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.999403731136737e-05,
+      "loss": 1.9402,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9994016591682057e-05,
+      "loss": 2.1131,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9993995836070618e-05,
+      "loss": 2.0474,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9993975044533127e-05,
+      "loss": 2.3384,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9993954217069658e-05,
+      "loss": 2.1698,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.999393335368029e-05,
+      "loss": 1.4312,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9993912454365093e-05,
+      "loss": 2.1241,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9993891519124142e-05,
+      "loss": 1.6609,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9993870547957516e-05,
+      "loss": 1.7796,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.999384954086529e-05,
+      "loss": 1.9498,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.999382849784754e-05,
+      "loss": 1.9202,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9993807418904336e-05,
+      "loss": 2.2076,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.999378630403576e-05,
+      "loss": 1.7091,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9993765153241886e-05,
+      "loss": 1.8078,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.999374396652279e-05,
+      "loss": 1.9429,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9993722743878546e-05,
+      "loss": 2.1163,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9993701485309232e-05,
+      "loss": 1.8052,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9993680190814927e-05,
+      "loss": 1.7507,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9993658860395705e-05,
+      "loss": 2.1057,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.999363749405164e-05,
+      "loss": 1.8303,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9993616091782818e-05,
+      "loss": 2.0109,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9993594653589302e-05,
+      "loss": 1.7804,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.999357317947118e-05,
+      "loss": 1.9734,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9993551669428526e-05,
+      "loss": 1.9686,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9993530123461418e-05,
+      "loss": 2.2621,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.999350854156993e-05,
+      "loss": 1.8942,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9993486923754145e-05,
+      "loss": 2.0255,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9993465270014132e-05,
+      "loss": 2.0386,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.9993443580349982e-05,
+      "loss": 1.7883,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9993421854761762e-05,
+      "loss": 1.7798,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9993400093249556e-05,
+      "loss": 2.0901,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.999337829581344e-05,
+      "loss": 1.7026,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.999335646245349e-05,
+      "loss": 1.9701,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9993334593169787e-05,
+      "loss": 1.6938,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9993312687962413e-05,
+      "loss": 2.0507,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.999329074683144e-05,
+      "loss": 1.8335,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.999326876977695e-05,
+      "loss": 1.6877,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.999324675679902e-05,
+      "loss": 1.8576,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.999322470789774e-05,
+      "loss": 1.9523,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9993202623073173e-05,
+      "loss": 2.0945,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9993180502325407e-05,
+      "loss": 1.8116,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.999315834565452e-05,
+      "loss": 2.4062,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9993136153060594e-05,
+      "loss": 1.9099,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9993113924543705e-05,
+      "loss": 2.0043,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.999309166010394e-05,
+      "loss": 1.9642,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9993069359741366e-05,
+      "loss": 1.8356,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9993047023456075e-05,
+      "loss": 1.9026,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.999302465124814e-05,
+      "loss": 2.0871,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9993002243117648e-05,
+      "loss": 1.8439,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9992979799064674e-05,
+      "loss": 2.1719,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9992957319089304e-05,
+      "loss": 2.3796,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9992934803191616e-05,
+      "loss": 2.3453,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.999291225137169e-05,
+      "loss": 1.6381,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9992889663629605e-05,
+      "loss": 1.7072,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9992867039965446e-05,
+      "loss": 1.9441,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9992844380379294e-05,
+      "loss": 1.8829,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.999282168487123e-05,
+      "loss": 2.4608,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9992798953441333e-05,
+      "loss": 1.873,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.999277618608969e-05,
+      "loss": 1.7939,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.999275338281638e-05,
+      "loss": 1.8231,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.999273054362148e-05,
+      "loss": 2.1601,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9992707668505082e-05,
+      "loss": 2.0867,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.999268475746726e-05,
+      "loss": 1.8325,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.99926618105081e-05,
+      "loss": 1.9365,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9992638827627686e-05,
+      "loss": 1.8011,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.99926158088261e-05,
+      "loss": 1.7284,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9992592754103418e-05,
+      "loss": 2.1529,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9992569663459733e-05,
+      "loss": 2.0192,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.999254653689512e-05,
+      "loss": 1.95,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9992523374409666e-05,
+      "loss": 1.8385,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9992500176003457e-05,
+      "loss": 1.7664,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.999247694167657e-05,
+      "loss": 2.0482,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.999245367142909e-05,
+      "loss": 1.74,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9992430365261106e-05,
+      "loss": 1.862,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.999240702317269e-05,
+      "loss": 2.1055,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9992383645163944e-05,
+      "loss": 2.1831,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.999236023123494e-05,
+      "loss": 1.9797,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.999233678138576e-05,
+      "loss": 1.7811,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.999231329561649e-05,
+      "loss": 2.1142,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9992289773927226e-05,
+      "loss": 1.9144,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9992266216318037e-05,
+      "loss": 1.6073,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9992242622789014e-05,
+      "loss": 1.8801,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9992218993340248e-05,
+      "loss": 2.1671,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9992195327971812e-05,
+      "loss": 2.0679,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9992171626683796e-05,
+      "loss": 1.8832,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.999214788947629e-05,
+      "loss": 2.0396,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9992124116349374e-05,
+      "loss": 1.8679,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9992100307303134e-05,
+      "loss": 2.0561,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9992076462337656e-05,
+      "loss": 1.8487,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9992052581453024e-05,
+      "loss": 2.0872,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9992028664649333e-05,
+      "loss": 2.1234,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9992004711926657e-05,
+      "loss": 1.9373,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.999198072328509e-05,
+      "loss": 2.2812,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9991956698724713e-05,
+      "loss": 1.892,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9991932638245616e-05,
+      "loss": 1.6789,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.999190854184788e-05,
+      "loss": 1.9232,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.99918844095316e-05,
+      "loss": 1.9962,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999186024129686e-05,
+      "loss": 2.0224,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9991836037143742e-05,
+      "loss": 1.8708,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.999181179707234e-05,
+      "loss": 2.3284,
+      "step": 685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9991787521082734e-05,
+      "loss": 1.8897,
+      "step": 686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9991763209175016e-05,
+      "loss": 1.7936,
+      "step": 687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9991738861349272e-05,
+      "loss": 2.2995,
+      "step": 688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.999171447760559e-05,
+      "loss": 1.773,
+      "step": 689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9991690057944057e-05,
+      "loss": 1.7802,
+      "step": 690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.999166560236476e-05,
+      "loss": 1.8885,
+      "step": 691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9991641110867792e-05,
+      "loss": 1.8498,
+      "step": 692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9991616583453235e-05,
+      "loss": 2.0115,
+      "step": 693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.999159202012118e-05,
+      "loss": 2.1106,
+      "step": 694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9991567420871715e-05,
+      "loss": 2.0506,
+      "step": 695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9991542785704924e-05,
+      "loss": 1.901,
+      "step": 696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9991518114620904e-05,
+      "loss": 1.9507,
+      "step": 697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.999149340761974e-05,
+      "loss": 2.0552,
+      "step": 698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9991468664701515e-05,
+      "loss": 1.9813,
+      "step": 699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.999144388586633e-05,
+      "loss": 1.7802,
+      "step": 700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.999141907111426e-05,
+      "loss": 1.7445,
+      "step": 701
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9991394220445407e-05,
+      "loss": 1.8601,
+      "step": 702
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.999136933385985e-05,
+      "loss": 1.9838,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9991344411357688e-05,
+      "loss": 2.1711,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9991319452939e-05,
+      "loss": 1.795,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9991294458603888e-05,
+      "loss": 1.7198,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.999126942835243e-05,
+      "loss": 2.1534,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9991244362184722e-05,
+      "loss": 1.9351,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9991219260100858e-05,
+      "loss": 1.9832,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.999119412210092e-05,
+      "loss": 1.8957,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9991168948185004e-05,
+      "loss": 1.9673,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9991143738353192e-05,
+      "loss": 1.9491,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.999111849260559e-05,
+      "loss": 2.012,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9991093210942273e-05,
+      "loss": 1.8603,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.999106789336334e-05,
+      "loss": 2.1721,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9991042539868887e-05,
+      "loss": 1.844,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.999101715045899e-05,
+      "loss": 1.7492,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9990991725133753e-05,
+      "loss": 2.0431,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9990966263893266e-05,
+      "loss": 2.1594,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9990940766737617e-05,
+      "loss": 1.9353,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9990915233666896e-05,
+      "loss": 1.7153,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.99908896646812e-05,
+      "loss": 1.8995,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9990864059780618e-05,
+      "loss": 1.9575,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25,
+      "learning_rate": 1.999083841896524e-05,
+      "loss": 1.6949,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9990812742235163e-05,
+      "loss": 2.0994,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9990787029590476e-05,
+      "loss": 1.7081,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9990761281031273e-05,
+      "loss": 1.9144,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9990735496557647e-05,
+      "loss": 2.1107,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9990709676169688e-05,
+      "loss": 1.8962,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9990683819867488e-05,
+      "loss": 1.9573,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.999065792765115e-05,
+      "loss": 1.865,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9990631999520754e-05,
+      "loss": 2.0414,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9990606035476397e-05,
+      "loss": 1.9226,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.999058003551818e-05,
+      "loss": 1.9187,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9990553999646186e-05,
+      "loss": 2.1443,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9990527927860516e-05,
+      "loss": 1.8235,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9990501820161254e-05,
+      "loss": 2.2828,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9990475676548508e-05,
+      "loss": 2.3617,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.999044949702236e-05,
+      "loss": 1.924,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.999042328158291e-05,
+      "loss": 1.917,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9990397030230256e-05,
+      "loss": 1.8803,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9990370742964483e-05,
+      "loss": 2.0377,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.999034441978569e-05,
+      "loss": 1.9628,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9990318060693973e-05,
+      "loss": 1.6645,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.999029166568942e-05,
+      "loss": 1.7766,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9990265234772134e-05,
+      "loss": 1.9625,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.999023876794221e-05,
+      "loss": 1.8082,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9990212265199738e-05,
+      "loss": 2.0716,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9990185726544814e-05,
+      "loss": 1.9333,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9990159151977536e-05,
+      "loss": 1.866,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9990132541498003e-05,
+      "loss": 1.9221,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.99901058951063e-05,
+      "loss": 1.7973,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.999007921280253e-05,
+      "loss": 1.9024,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9990052494586792e-05,
+      "loss": 3.6333,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9990025740459172e-05,
+      "loss": 2.0112,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9989998950419774e-05,
+      "loss": 2.0302,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9989972124468697e-05,
+      "loss": 1.8621,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9989945262606024e-05,
+      "loss": 1.8082,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.998991836483187e-05,
+      "loss": 1.9573,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9989891431146315e-05,
+      "loss": 2.125,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9989864461549464e-05,
+      "loss": 2.072,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9989837456041418e-05,
+      "loss": 1.8101,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9989810414622265e-05,
+      "loss": 1.6974,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9989783337292104e-05,
+      "loss": 1.7939,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9989756224051035e-05,
+      "loss": 2.0026,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9989729074899155e-05,
+      "loss": 1.9085,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9989701889836563e-05,
+      "loss": 1.9374,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9989674668863356e-05,
+      "loss": 2.2994,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9989647411979632e-05,
+      "loss": 1.9762,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9989620119185485e-05,
+      "loss": 2.2087,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9989592790481017e-05,
+      "loss": 2.1901,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9989565425866323e-05,
+      "loss": 2.248,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9989538025341508e-05,
+      "loss": 2.1288,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9989510588906663e-05,
+      "loss": 2.1048,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9989483116561894e-05,
+      "loss": 1.9344,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.998945560830729e-05,
+      "loss": 1.9392,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.998942806414296e-05,
+      "loss": 1.6935,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9989400484068998e-05,
+      "loss": 1.9068,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9989372868085503e-05,
+      "loss": 1.7771,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9989345216192575e-05,
+      "loss": 1.7333,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.998931752839031e-05,
+      "loss": 1.7531,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9989289804678815e-05,
+      "loss": 1.8865,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9989262045058187e-05,
+      "loss": 2.0736,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.998923424952852e-05,
+      "loss": 1.792,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.998920641808992e-05,
+      "loss": 1.7203,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9989178550742484e-05,
+      "loss": 2.1117,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9989150647486318e-05,
+      "loss": 1.7714,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9989122708321512e-05,
+      "loss": 1.8845,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9989094733248177e-05,
+      "loss": 2.0226,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9989066722266405e-05,
+      "loss": 1.7316,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9989038675376305e-05,
+      "loss": 1.9118,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.998901059257797e-05,
+      "loss": 1.8748,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9988982473871507e-05,
+      "loss": 2.524,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.998895431925701e-05,
+      "loss": 2.0812,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9988926128734583e-05,
+      "loss": 2.1018,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9988897902304337e-05,
+      "loss": 2.2356,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.998886963996636e-05,
+      "loss": 1.9,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9988841341720758e-05,
+      "loss": 1.7283,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9988813007567636e-05,
+      "loss": 1.7364,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9988784637507093e-05,
+      "loss": 2.0212,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.998875623153923e-05,
+      "loss": 1.7975,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9988727789664156e-05,
+      "loss": 2.1516,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9988699311881964e-05,
+      "loss": 1.9888,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.998867079819276e-05,
+      "loss": 1.4943,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9988642248596647e-05,
+      "loss": 2.304,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9988613663093725e-05,
+      "loss": 1.5896,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9988585041684103e-05,
+      "loss": 2.0471,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9988556384367878e-05,
+      "loss": 1.9205,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9988527691145157e-05,
+      "loss": 1.7027,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9988498962016038e-05,
+      "loss": 1.9795,
+      "step": 810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.998847019698063e-05,
+      "loss": 2.1078,
+      "step": 811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9988441396039032e-05,
+      "loss": 2.1084,
+      "step": 812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.998841255919135e-05,
+      "loss": 2.1076,
+      "step": 813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.998838368643769e-05,
+      "loss": 1.9984,
+      "step": 814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.998835477777815e-05,
+      "loss": 1.8957,
+      "step": 815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9988325833212837e-05,
+      "loss": 2.032,
+      "step": 816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9988296852741853e-05,
+      "loss": 1.9019,
+      "step": 817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.998826783636531e-05,
+      "loss": 1.8987,
+      "step": 818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9988238784083302e-05,
+      "loss": 1.8819,
+      "step": 819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9988209695895934e-05,
+      "loss": 1.8925,
+      "step": 820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9988180571803318e-05,
+      "loss": 2.0545,
+      "step": 821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9988151411805554e-05,
+      "loss": 1.7171,
+      "step": 822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.998812221590275e-05,
+      "loss": 1.9133,
+      "step": 823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.998809298409501e-05,
+      "loss": 2.0934,
+      "step": 824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9988063716382433e-05,
+      "loss": 1.9704,
+      "step": 825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.998803441276513e-05,
+      "loss": 1.8664,
+      "step": 826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.998800507324321e-05,
+      "loss": 2.1401,
+      "step": 827
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9987975697816768e-05,
+      "loss": 2.0238,
+      "step": 828
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.998794628648592e-05,
+      "loss": 1.7241,
+      "step": 829
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9987916839250764e-05,
+      "loss": 2.2265,
+      "step": 830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9987887356111414e-05,
+      "loss": 2.1353,
+      "step": 831
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9987857837067965e-05,
+      "loss": 1.8933,
+      "step": 832
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.998782828212053e-05,
+      "loss": 1.8862,
+      "step": 833
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.998779869126922e-05,
+      "loss": 2.1843,
+      "step": 834
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.998776906451413e-05,
+      "loss": 2.141,
+      "step": 835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.9987739401855376e-05,
+      "loss": 0.7857,
+      "step": 836
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.998770970329306e-05,
+      "loss": 1.8143,
+      "step": 837
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9987679968827284e-05,
+      "loss": 1.9695,
+      "step": 838
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9987650198458168e-05,
+      "loss": 2.0701,
+      "step": 839
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9987620392185805e-05,
+      "loss": 2.0536,
+      "step": 840
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9987590550010315e-05,
+      "loss": 1.6851,
+      "step": 841
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9987560671931795e-05,
+      "loss": 1.7336,
+      "step": 842
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9987530757950357e-05,
+      "loss": 1.9653,
+      "step": 843
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.998750080806611e-05,
+      "loss": 1.8866,
+      "step": 844
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9987470822279157e-05,
+      "loss": 1.7826,
+      "step": 845
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.998744080058961e-05,
+      "loss": 1.9681,
+      "step": 846
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9987410742997574e-05,
+      "loss": 2.2377,
+      "step": 847
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.998738064950316e-05,
+      "loss": 1.7823,
+      "step": 848
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9987350520106474e-05,
+      "loss": 1.6276,
+      "step": 849
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9987320354807623e-05,
+      "loss": 2.062,
+      "step": 850
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.998729015360672e-05,
+      "loss": 1.9354,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.9987259916503868e-05,
+      "loss": 1.6743,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9987229643499184e-05,
+      "loss": 1.799,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9987199334592764e-05,
+      "loss": 1.8388,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9987168989784728e-05,
+      "loss": 1.8083,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.998713860907518e-05,
+      "loss": 2.0398,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.998710819246423e-05,
+      "loss": 1.7847,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.998707773995199e-05,
+      "loss": 2.0118,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.998704725153857e-05,
+      "loss": 1.8771,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9987016727224075e-05,
+      "loss": 1.7617,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9986986167008614e-05,
+      "loss": 2.0752,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9986955570892302e-05,
+      "loss": 2.1463,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9986924938875245e-05,
+      "loss": 2.27,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9986894270957556e-05,
+      "loss": 1.9798,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.998686356713934e-05,
+      "loss": 2.1508,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.998683282742072e-05,
+      "loss": 1.8692,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.998680205180179e-05,
+      "loss": 1.9352,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.998677124028267e-05,
+      "loss": 1.8193,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9986740392863472e-05,
+      "loss": 1.9231,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9986709509544302e-05,
+      "loss": 1.6526,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9986678590325273e-05,
+      "loss": 1.8691,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.99866476352065e-05,
+      "loss": 2.3695,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9986616644188085e-05,
+      "loss": 1.9907,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9986585617270146e-05,
+      "loss": 1.9106,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9986554554452795e-05,
+      "loss": 1.8539,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9986523455736142e-05,
+      "loss": 2.1082,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9986492321120298e-05,
+      "loss": 1.9737,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9986461150605378e-05,
+      "loss": 2.0784,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9986429944191487e-05,
+      "loss": 2.32,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9986398701878748e-05,
+      "loss": 1.7908,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9986367423667264e-05,
+      "loss": 2.0033,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9986336109557153e-05,
+      "loss": 1.9085,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.998630475954852e-05,
+      "loss": 1.7585,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.998627337364149e-05,
+      "loss": 1.8354,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9986241951836165e-05,
+      "loss": 1.8307,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9986210494132664e-05,
+      "loss": 2.0407,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9986179000531098e-05,
+      "loss": 1.9755,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9986147471031576e-05,
+      "loss": 2.0469,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.998611590563422e-05,
+      "loss": 1.9214,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9986084304339136e-05,
+      "loss": 1.7955,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9986052667146443e-05,
+      "loss": 2.0925,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9986020994056253e-05,
+      "loss": 2.1088,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9985989285068675e-05,
+      "loss": 2.167,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.998595754018383e-05,
+      "loss": 1.6543,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.998592575940183e-05,
+      "loss": 1.5747,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9985893942722787e-05,
+      "loss": 1.8589,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9985862090146817e-05,
+      "loss": 1.7167,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9985830201674035e-05,
+      "loss": 1.8394,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9985798277304552e-05,
+      "loss": 1.5749,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.998576631703849e-05,
+      "loss": 2.0822,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.998573432087596e-05,
+      "loss": 2.1473,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9985702288817076e-05,
+      "loss": 1.9667,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.998567022086195e-05,
+      "loss": 1.8481,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9985638117010705e-05,
+      "loss": 1.7084,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.998560597726345e-05,
+      "loss": 1.9072,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9985573801620302e-05,
+      "loss": 1.9388,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9985541590081382e-05,
+      "loss": 1.8124,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9985509342646797e-05,
+      "loss": 1.7247,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9985477059316675e-05,
+      "loss": 2.0208,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9985444740091115e-05,
+      "loss": 1.5908,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.998541238497025e-05,
+      "loss": 2.4039,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9985379993954182e-05,
+      "loss": 2.1507,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.998534756704304e-05,
+      "loss": 2.0267,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9985315104236936e-05,
+      "loss": 2.2359,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9985282605535984e-05,
+      "loss": 1.7535,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.99852500709403e-05,
+      "loss": 2.0774,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9985217500450002e-05,
+      "loss": 1.9101,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9985184894065213e-05,
+      "loss": 2.2844,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9985152251786044e-05,
+      "loss": 1.9977,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9985119573612614e-05,
+      "loss": 1.6968,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.998508685954504e-05,
+      "loss": 2.058,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.998505410958344e-05,
+      "loss": 2.0004,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9985021323727935e-05,
+      "loss": 1.9341,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9984988501978636e-05,
+      "loss": 2.0332,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.998495564433567e-05,
+      "loss": 2.0575,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.998492275079914e-05,
+      "loss": 2.1987,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9984889821369183e-05,
+      "loss": 1.9363,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9984856856045905e-05,
+      "loss": 2.0041,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.998482385482943e-05,
+      "loss": 1.9,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9984790817719873e-05,
+      "loss": 1.7162,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9984757744717352e-05,
+      "loss": 2.0255,
+      "step": 931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9984724635821993e-05,
+      "loss": 1.8735,
+      "step": 932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9984691491033908e-05,
+      "loss": 1.6875,
+      "step": 933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9984658310353218e-05,
+      "loss": 1.9192,
+      "step": 934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9984625093780042e-05,
+      "loss": 1.7047,
+      "step": 935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9984591841314503e-05,
+      "loss": 1.9788,
+      "step": 936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9984558552956715e-05,
+      "loss": 1.8083,
+      "step": 937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.75,
+      "learning_rate": 1.99845252287068e-05,
+      "loss": 2.0561,
+      "step": 938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.998449186856488e-05,
+      "loss": 1.7959,
+      "step": 939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9984458472531072e-05,
+      "loss": 1.8278,
+      "step": 940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.99844250406055e-05,
+      "loss": 2.0424,
+      "step": 941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9984391572788277e-05,
+      "loss": 1.8588,
+      "step": 942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9984358069079528e-05,
+      "loss": 1.8278,
+      "step": 943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9984324529479374e-05,
+      "loss": 1.6852,
+      "step": 944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9984290953987933e-05,
+      "loss": 2.5038,
+      "step": 945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.998425734260533e-05,
+      "loss": 1.8993,
+      "step": 946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.75,
+      "learning_rate": 1.9984223695331678e-05,
+      "loss": 2.3728,
+      "step": 947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.998419001216711e-05,
+      "loss": 1.9324,
+      "step": 948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9984156293111737e-05,
+      "loss": 2.1019,
+      "step": 949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.9984122538165684e-05,
+      "loss": 2.0973,
+      "step": 950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.998408874732907e-05,
+      "loss": 2.2412,
+      "step": 951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.998405492060202e-05,
+      "loss": 2.1486,
+      "step": 952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.998402105798465e-05,
+      "loss": 2.0294,
+      "step": 953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9983987159477088e-05,
+      "loss": 2.1896,
+      "step": 954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9983953225079455e-05,
+      "loss": 1.9844,
+      "step": 955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.998391925479187e-05,
+      "loss": 2.0899,
+      "step": 956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9983885248614456e-05,
+      "loss": 1.8506,
+      "step": 957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9983851206547334e-05,
+      "loss": 1.8567,
+      "step": 958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.998381712859063e-05,
+      "loss": 1.7506,
+      "step": 959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9983783014744465e-05,
+      "loss": 2.0465,
+      "step": 960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.998374886500896e-05,
+      "loss": 1.9238,
+      "step": 961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9983714679384237e-05,
+      "loss": 1.7657,
+      "step": 962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9983680457870424e-05,
+      "loss": 1.8991,
+      "step": 963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9983646200467642e-05,
+      "loss": 1.6966,
+      "step": 964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.998361190717601e-05,
+      "loss": 1.9527,
+      "step": 965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9983577577995656e-05,
+      "loss": 1.8751,
+      "step": 966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9983543212926702e-05,
+      "loss": 1.9911,
+      "step": 967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.998350881196927e-05,
+      "loss": 1.9687,
+      "step": 968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9983474375123484e-05,
+      "loss": 1.8365,
+      "step": 969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.998343990238947e-05,
+      "loss": 1.9128,
+      "step": 970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.998340539376735e-05,
+      "loss": 2.0355,
+      "step": 971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.998337084925725e-05,
+      "loss": 1.853,
+      "step": 972
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.998333626885929e-05,
+      "loss": 1.6208,
+      "step": 973
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9983301652573598e-05,
+      "loss": 1.9626,
+      "step": 974
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.99832670004003e-05,
+      "loss": 2.1263,
+      "step": 975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9983232312339516e-05,
+      "loss": 2.0663,
+      "step": 976
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9983197588391373e-05,
+      "loss": 1.646,
+      "step": 977
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9983162828555996e-05,
+      "loss": 1.8988,
+      "step": 978
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.998312803283351e-05,
+      "loss": 1.9036,
+      "step": 979
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9983093201224036e-05,
+      "loss": 2.054,
+      "step": 980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.998305833372771e-05,
+      "loss": 1.6617,
+      "step": 981
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9983023430344645e-05,
+      "loss": 2.2001,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.998298849107497e-05,
+      "loss": 1.9852,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9982953515918816e-05,
+      "loss": 1.873,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9982918504876306e-05,
+      "loss": 1.7489,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.998288345794756e-05,
+      "loss": 1.9994,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9982848375132714e-05,
+      "loss": 2.0431,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9982813256431886e-05,
+      "loss": 1.974,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9982778101845206e-05,
+      "loss": 1.9944,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.99827429113728e-05,
+      "loss": 1.7623,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9982707685014793e-05,
+      "loss": 1.8842,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9982672422771313e-05,
+      "loss": 1.6115,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.998263712464249e-05,
+      "loss": 1.699,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.998260179062844e-05,
+      "loss": 1.8745,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9982566420729303e-05,
+      "loss": 2.0371,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.99825310149452e-05,
+      "loss": 1.7351,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9982495573276256e-05,
+      "loss": 1.7945,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9982460095722605e-05,
+      "loss": 1.4957,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.998242458228437e-05,
+      "loss": 2.0088,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9982389032961674e-05,
+      "loss": 1.6381,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.998235344775466e-05,
+      "loss": 1.9408,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9982317826663438e-05,
+      "loss": 1.8522,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9982282169688147e-05,
+      "loss": 2.1761,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.998224647682891e-05,
+      "loss": 1.8544,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9982210748085862e-05,
+      "loss": 1.8067,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9982174983459125e-05,
+      "loss": 1.9111,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9982139182948827e-05,
+      "loss": 1.9748,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9982103346555105e-05,
+      "loss": 1.906,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9982067474278077e-05,
+      "loss": 2.2672,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.998203156611788e-05,
+      "loss": 1.8788,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.998199562207464e-05,
+      "loss": 1.891,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9981959642148485e-05,
+      "loss": 2.1624,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9981923626339545e-05,
+      "loss": 2.0586,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9981887574647955e-05,
+      "loss": 2.4988,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9981851487073833e-05,
+      "loss": 1.7894,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.998181536361732e-05,
+      "loss": 1.755,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.998177920427854e-05,
+      "loss": 1.9659,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9981743009057625e-05,
+      "loss": 2.0868,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9981706777954705e-05,
+      "loss": 2.08,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9981670510969908e-05,
+      "loss": 1.8135,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.998163420810337e-05,
+      "loss": 2.3319,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9981597869355213e-05,
+      "loss": 1.76,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9981561494725573e-05,
+      "loss": 1.9674,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.998152508421458e-05,
+      "loss": 1.8395,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9981488637822368e-05,
+      "loss": 2.1975,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9981452155549063e-05,
+      "loss": 2.1846,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.99814156373948e-05,
+      "loss": 1.9574,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9981379083359705e-05,
+      "loss": 2.2409,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9981342493443915e-05,
+      "loss": 1.7956,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.998130586764756e-05,
+      "loss": 2.049,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9981269205970765e-05,
+      "loss": 2.4745,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9981232508413673e-05,
+      "loss": 1.9379,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.998119577497641e-05,
+      "loss": 1.9489,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9981159005659107e-05,
+      "loss": 1.9889,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9981122200461898e-05,
+      "loss": 1.6175,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.998108535938492e-05,
+      "loss": 1.8249,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9981048482428295e-05,
+      "loss": 1.4871,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.998101156959216e-05,
+      "loss": 2.0047,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.998097462087665e-05,
+      "loss": 1.9787,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9980937636281902e-05,
+      "loss": 1.9377,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.998090061580804e-05,
+      "loss": 1.9925,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.99808635594552e-05,
+      "loss": 1.845,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9980826467223517e-05,
+      "loss": 1.7766,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9980789339113124e-05,
+      "loss": 2.2217,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9980752175124152e-05,
+      "loss": 1.832,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9980714975256735e-05,
+      "loss": 2.0915,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9980677739511012e-05,
+      "loss": 1.9409,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.998064046788711e-05,
+      "loss": 1.6934,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9980603160385165e-05,
+      "loss": 1.8563,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9980565817005315e-05,
+      "loss": 1.8503,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.998052843774769e-05,
+      "loss": 1.7233,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9980491022612427e-05,
+      "loss": 1.7077,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9980453571599657e-05,
+      "loss": 1.8954,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.375,
+      "learning_rate": 1.998041608470952e-05,
+      "loss": 1.5666,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.998037856194214e-05,
+      "loss": 1.8931,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9980341003297667e-05,
+      "loss": 2.0165,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9980303408776224e-05,
+      "loss": 1.7443,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9980265778377956e-05,
+      "loss": 1.9838,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9980228112102987e-05,
+      "loss": 2.0502,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.998019040995146e-05,
+      "loss": 1.867,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9980152671923506e-05,
+      "loss": 1.8295,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9980114898019266e-05,
+      "loss": 1.7923,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9980077088238877e-05,
+      "loss": 2.0772,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9980039242582466e-05,
+      "loss": 1.8865,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9980001361050173e-05,
+      "loss": 1.7207,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.997996344364214e-05,
+      "loss": 2.441,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9979925490358496e-05,
+      "loss": 1.66,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9979887501199382e-05,
+      "loss": 1.594,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.997984947616493e-05,
+      "loss": 2.0439,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.997981141525528e-05,
+      "loss": 1.9754,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.997977331847057e-05,
+      "loss": 1.9697,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9979735185810934e-05,
+      "loss": 1.7778,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9979697017276507e-05,
+      "loss": 1.8499,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9979658812867436e-05,
+      "loss": 2.01,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9979620572583846e-05,
+      "loss": 1.669,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.997958229642588e-05,
+      "loss": 1.945,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.997954398439368e-05,
+      "loss": 1.7965,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9979505636487375e-05,
+      "loss": 1.6025,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.997946725270711e-05,
+      "loss": 1.8061,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.997942883305302e-05,
+      "loss": 2.004,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9979390377525244e-05,
+      "loss": 1.8236,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9979351886123917e-05,
+      "loss": 1.8455,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9979313358849183e-05,
+      "loss": 2.0813,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9979274795701178e-05,
+      "loss": 2.1259,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25,
+      "learning_rate": 1.997923619668004e-05,
+      "loss": 1.6095,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9979197561785907e-05,
+      "loss": 1.8083,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9979158891018916e-05,
+      "loss": 2.0633,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.997912018437921e-05,
+      "loss": 1.9851,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.997908144186693e-05,
+      "loss": 2.0316,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.997904266348221e-05,
+      "loss": 2.1711,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.997900384922519e-05,
+      "loss": 1.7913,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9978964999096013e-05,
+      "loss": 2.0605,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9978926113094816e-05,
+      "loss": 1.9017,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.997888719122174e-05,
+      "loss": 2.078,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9978848233476928e-05,
+      "loss": 1.9626,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9978809239860514e-05,
+      "loss": 1.7996,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9978770210372637e-05,
+      "loss": 1.8281,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9978731145013444e-05,
+      "loss": 1.8686,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9978692043783073e-05,
+      "loss": 1.8907,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9978652906681662e-05,
+      "loss": 1.8932,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9978613733709356e-05,
+      "loss": 1.9688,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.997857452486629e-05,
+      "loss": 1.8565,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9978535280152614e-05,
+      "loss": 1.9988,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.997849599956846e-05,
+      "loss": 1.8032,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9978456683113974e-05,
+      "loss": 2.0094,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9978417330789295e-05,
+      "loss": 1.855,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9978377942594566e-05,
+      "loss": 1.9888,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9978338518529927e-05,
+      "loss": 1.9761,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9978299058595522e-05,
+      "loss": 1.7885,
+      "step": 1109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.997825956279149e-05,
+      "loss": 2.2883,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9978220031117977e-05,
+      "loss": 2.0935,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9978180463575125e-05,
+      "loss": 1.7272,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.997814086016307e-05,
+      "loss": 2.0237,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.997810122088196e-05,
+      "loss": 2.1199,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.997806154573193e-05,
+      "loss": 2.1757,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9978021834713136e-05,
+      "loss": 1.8424,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.9977982087825714e-05,
+      "loss": 2.3412,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.99779423050698e-05,
+      "loss": 1.9278,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9977902486445546e-05,
+      "loss": 1.8757,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9977862631953094e-05,
+      "loss": 1.7203,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9977822741592586e-05,
+      "loss": 2.1431,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25,
+      "learning_rate": 1.997778281536416e-05,
+      "loss": 1.8565,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.997774285326797e-05,
+      "loss": 2.0002,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9977702855304152e-05,
+      "loss": 1.9959,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.997766282147285e-05,
+      "loss": 2.0089,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.997762275177421e-05,
+      "loss": 2.2061,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.997758264620838e-05,
+      "loss": 1.6958,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9977542504775497e-05,
+      "loss": 1.6767,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9977502327475706e-05,
+      "loss": 1.9492,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9977462114309155e-05,
+      "loss": 1.7988,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.997742186527599e-05,
+      "loss": 1.8311,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.997738158037635e-05,
+      "loss": 1.8348,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9977341259610385e-05,
+      "loss": 1.7568,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.997730090297824e-05,
+      "loss": 2.3445,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9977260510480052e-05,
+      "loss": 1.7815,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9977220082115972e-05,
+      "loss": 1.7121,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9977179617886152e-05,
+      "loss": 1.6129,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9977139117790725e-05,
+      "loss": 2.3378,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9977098581829842e-05,
+      "loss": 1.7274,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9977058010003656e-05,
+      "loss": 2.12,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.99770174023123e-05,
+      "loss": 1.806,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9976976758755925e-05,
+      "loss": 1.7689,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.997693607933468e-05,
+      "loss": 2.0757,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.997689536404871e-05,
+      "loss": 1.7139,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9976854612898163e-05,
+      "loss": 1.8994,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9976813825883182e-05,
+      "loss": 1.6997,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9976773003003912e-05,
+      "loss": 1.8706,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.99767321442605e-05,
+      "loss": 1.9195,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9976691249653102e-05,
+      "loss": 1.9474,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9976650319181857e-05,
+      "loss": 1.8497,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9976609352846908e-05,
+      "loss": 1.6431,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9976568350648415e-05,
+      "loss": 1.8532,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.9976527312586516e-05,
+      "loss": 2.1058,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9976486238661358e-05,
+      "loss": 1.7971,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9976445128873092e-05,
+      "loss": 1.8243,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9976403983221865e-05,
+      "loss": 2.1172,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9976362801707828e-05,
+      "loss": 1.7304,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9976321584331125e-05,
+      "loss": 2.0576,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9976280331091903e-05,
+      "loss": 2.1057,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9976239041990314e-05,
+      "loss": 2.0001,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9976197717026507e-05,
+      "loss": 2.006,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9976156356200624e-05,
+      "loss": 1.8633,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.997611495951282e-05,
+      "loss": 1.7895,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.997607352696324e-05,
+      "loss": 1.9987,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.997603205855204e-05,
+      "loss": 1.8034,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9975990554279363e-05,
+      "loss": 1.457,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9975949014145356e-05,
+      "loss": 1.8031,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9975907438150176e-05,
+      "loss": 1.9038,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9975865826293966e-05,
+      "loss": 1.86,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9975824178576875e-05,
+      "loss": 1.849,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9975782494999058e-05,
+      "loss": 1.8096,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9975740775560658e-05,
+      "loss": 2.5804,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9975699020261838e-05,
+      "loss": 1.8041,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.997565722910273e-05,
+      "loss": 1.9537,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9975615402083498e-05,
+      "loss": 1.7845,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9975573539204286e-05,
+      "loss": 1.824,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9975531640465247e-05,
+      "loss": 2.0434,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.997548970586653e-05,
+      "loss": 1.582,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.997544773540829e-05,
+      "loss": 1.8823,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9975405729090672e-05,
+      "loss": 2.088,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9975363686913827e-05,
+      "loss": 1.9862,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9975321608877914e-05,
+      "loss": 1.7335,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9975279494983077e-05,
+      "loss": 2.0799,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9975237345229466e-05,
+      "loss": 1.9892,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.997519515961724e-05,
+      "loss": 2.1355,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9975152938146543e-05,
+      "loss": 2.1918,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.9975110680817533e-05,
+      "loss": 1.999,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9975068387630358e-05,
+      "loss": 1.8594,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.997502605858517e-05,
+      "loss": 1.7278,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9974983693682126e-05,
+      "loss": 2.0346,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9974941292921374e-05,
+      "loss": 1.9226,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9974898856303064e-05,
+      "loss": 1.8672,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9974856383827356e-05,
+      "loss": 2.0116,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9974813875494392e-05,
+      "loss": 2.0784,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9974771331304335e-05,
+      "loss": 1.8868,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9974728751257336e-05,
+      "loss": 1.7517,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.9974686135353543e-05,
+      "loss": 1.7851,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9974643483593113e-05,
+      "loss": 2.0336,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9974600795976197e-05,
+      "loss": 1.9843,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9974558072502953e-05,
+      "loss": 1.5982,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9974515313173526e-05,
+      "loss": 2.3631,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.997447251798808e-05,
+      "loss": 1.7583,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9974429686946765e-05,
+      "loss": 1.9419,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9974386820049735e-05,
+      "loss": 1.8526,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.997434391729714e-05,
+      "loss": 1.9831,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9974300978689138e-05,
+      "loss": 1.9337,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9974258004225883e-05,
+      "loss": 1.6927,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.997421499390753e-05,
+      "loss": 2.1581,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9974171947734235e-05,
+      "loss": 2.0997,
+      "step": 1209
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9974128865706147e-05,
+      "loss": 2.3669,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9974085747823425e-05,
+      "loss": 1.7894,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9974042594086226e-05,
+      "loss": 1.9011,
+      "step": 1212
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.99739994044947e-05,
+      "loss": 2.1271,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9973956179049008e-05,
+      "loss": 1.72,
+      "step": 1214
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.99739129177493e-05,
+      "loss": 2.0763,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9973869620595735e-05,
+      "loss": 1.947,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9973826287588465e-05,
+      "loss": 1.9477,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.997378291872765e-05,
+      "loss": 2.1404,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9973739514013445e-05,
+      "loss": 1.7101,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9973696073446004e-05,
+      "loss": 1.9941,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9973652597025484e-05,
+      "loss": 1.911,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9973609084752042e-05,
+      "loss": 2.0607,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9973565536625834e-05,
+      "loss": 1.9674,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9973521952647017e-05,
+      "loss": 1.923,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9973478332815746e-05,
+      "loss": 2.431,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9973434677132182e-05,
+      "loss": 1.8258,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9973390985596475e-05,
+      "loss": 1.9492,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.997334725820879e-05,
+      "loss": 1.6146,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.997330349496928e-05,
+      "loss": 2.0894,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9973259695878097e-05,
+      "loss": 2.016,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.997321586093541e-05,
+      "loss": 1.9777,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9973171990141367e-05,
+      "loss": 1.7788,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.997312808349613e-05,
+      "loss": 2.408,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.997308414099986e-05,
+      "loss": 1.5292,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9973040162652707e-05,
+      "loss": 1.919,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9972996148454836e-05,
+      "loss": 2.0356,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.99729520984064e-05,
+      "loss": 1.8954,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9972908012507558e-05,
+      "loss": 1.8517,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9972863890758474e-05,
+      "loss": 1.7124,
+      "step": 1239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.99728197331593e-05,
+      "loss": 1.7699,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9972775539710198e-05,
+      "loss": 1.8236,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.997273131041133e-05,
+      "loss": 2.2363,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.997268704526285e-05,
+      "loss": 2.1519,
+      "step": 1243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9972642744264914e-05,
+      "loss": 1.7978,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9972598407417688e-05,
+      "loss": 2.79,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.997255403472133e-05,
+      "loss": 1.621,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9972509626176e-05,
+      "loss": 1.8351,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9972465181781856e-05,
+      "loss": 2.1487,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.997242070153906e-05,
+      "loss": 1.6506,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9972376185447767e-05,
+      "loss": 1.9271,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.997233163350814e-05,
+      "loss": 1.926,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.997228704572034e-05,
+      "loss": 1.8001,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9972242422084524e-05,
+      "loss": 1.8906,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.997219776260086e-05,
+      "loss": 1.9607,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.99721530672695e-05,
+      "loss": 1.8354,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9972108336090605e-05,
+      "loss": 1.6412,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9972063569064344e-05,
+      "loss": 2.0008,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.997201876619087e-05,
+      "loss": 1.7729,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.997197392747035e-05,
+      "loss": 1.8392,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9971929052902937e-05,
+      "loss": 1.9749,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.99718841424888e-05,
+      "loss": 2.0559,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.99718391962281e-05,
+      "loss": 2.1253,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9971794214120996e-05,
+      "loss": 1.9318,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9971749196167648e-05,
+      "loss": 1.8877,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.997170414236822e-05,
+      "loss": 1.7133,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9971659052722873e-05,
+      "loss": 1.8839,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.375,
+      "learning_rate": 1.997161392723177e-05,
+      "loss": 1.6639,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9971568765895075e-05,
+      "loss": 1.8338,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9971523568712947e-05,
+      "loss": 2.1027,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9971478335685554e-05,
+      "loss": 2.0492,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9971433066813048e-05,
+      "loss": 2.0592,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9971387762095605e-05,
+      "loss": 2.0642,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9971342421533374e-05,
+      "loss": 1.5698,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.997129704512653e-05,
+      "loss": 2.0192,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.997125163287523e-05,
+      "loss": 2.0498,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.997120618477964e-05,
+      "loss": 1.7998,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.997116070083992e-05,
+      "loss": 1.779,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9971115181056237e-05,
+      "loss": 1.7969,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9971069625428753e-05,
+      "loss": 1.9623,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.997102403395763e-05,
+      "loss": 1.9546,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9970978406643034e-05,
+      "loss": 1.939,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.997093274348513e-05,
+      "loss": 1.909,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.997088704448408e-05,
+      "loss": 1.9609,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9970841309640047e-05,
+      "loss": 1.8303,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9970795538953202e-05,
+      "loss": 1.6812,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9970749732423703e-05,
+      "loss": 2.048,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9970703890051714e-05,
+      "loss": 1.8234,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9970658011837404e-05,
+      "loss": 2.098,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.997061209778094e-05,
+      "loss": 1.8938,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.997056614788248e-05,
+      "loss": 2.1119,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9970520162142194e-05,
+      "loss": 1.8267,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9970474140560246e-05,
+      "loss": 1.9373,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.99704280831368e-05,
+      "loss": 2.0156,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9970381989872026e-05,
+      "loss": 2.1155,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9970335860766082e-05,
+      "loss": 1.9589,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9970289695819142e-05,
+      "loss": 2.2975,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.997024349503137e-05,
+      "loss": 2.0978,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9970197258402927e-05,
+      "loss": 2.0272,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.997015098593398e-05,
+      "loss": 1.9376,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9970104677624706e-05,
+      "loss": 1.9988,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.997005833347526e-05,
+      "loss": 1.8336,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9970011953485812e-05,
+      "loss": 2.0024,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.996996553765653e-05,
+      "loss": 2.032,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9969919085987575e-05,
+      "loss": 1.927,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.996987259847912e-05,
+      "loss": 1.8857,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9969826075131332e-05,
+      "loss": 1.9016,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.996977951594438e-05,
+      "loss": 1.8754,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9969732920918426e-05,
+      "loss": 1.8991,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9969686290053644e-05,
+      "loss": 2.1741,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.996963962335019e-05,
+      "loss": 2.012,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9969592920808244e-05,
+      "loss": 1.9392,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9969546182427968e-05,
+      "loss": 1.9568,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.996949940820953e-05,
+      "loss": 1.7672,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.99694525981531e-05,
+      "loss": 1.9767,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9969405752258847e-05,
+      "loss": 1.9653,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9969358870526938e-05,
+      "loss": 2.0585,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9969311952957543e-05,
+      "loss": 1.2525,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9969264999550825e-05,
+      "loss": 1.6079,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9969218010306962e-05,
+      "loss": 1.8699,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.9969170985226115e-05,
+      "loss": 1.499,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9969123924308452e-05,
+      "loss": 2.1032,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9969076827554153e-05,
+      "loss": 2.2314,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9969029694963375e-05,
+      "loss": 2.1406,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.99689825265363e-05,
+      "loss": 1.7482,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9968935322273082e-05,
+      "loss": 1.8857,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9968888082173904e-05,
+      "loss": 1.7733,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9968840806238933e-05,
+      "loss": 1.8974,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.996879349446833e-05,
+      "loss": 2.0649,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9968746146862276e-05,
+      "loss": 1.9076,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9968698763420938e-05,
+      "loss": 1.9283,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9968651344144482e-05,
+      "loss": 1.8717,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9968603889033088e-05,
+      "loss": 2.1366,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9968556398086915e-05,
+      "loss": 2.1037,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.996850887130614e-05,
+      "loss": 2.1221,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9968461308690933e-05,
+      "loss": 1.606,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.996841371024147e-05,
+      "loss": 1.8627,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9968366075957914e-05,
+      "loss": 2.0843,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9968318405840437e-05,
+      "loss": 1.597,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.996827069988921e-05,
+      "loss": 1.8394,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9968222958104415e-05,
+      "loss": 1.9951,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9968175180486212e-05,
+      "loss": 1.7905,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.996812736703478e-05,
+      "loss": 1.8541,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9968079517750284e-05,
+      "loss": 2.1498,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.99680316326329e-05,
+      "loss": 2.122,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9967983711682797e-05,
+      "loss": 2.0242,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9967935754900155e-05,
+      "loss": 2.1699,
+      "step": 1346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.996788776228514e-05,
+      "loss": 2.078,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9967839733837923e-05,
+      "loss": 1.8037,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.996779166955868e-05,
+      "loss": 2.1516,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9967743569447585e-05,
+      "loss": 2.3139,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9967695433504813e-05,
+      "loss": 1.9256,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9967647261730528e-05,
+      "loss": 2.0519,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.996759905412491e-05,
+      "loss": 1.9297,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.996755081068813e-05,
+      "loss": 1.8785,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.996750253142036e-05,
+      "loss": 2.0829,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9967454216321778e-05,
+      "loss": 2.2776,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9967405865392554e-05,
+      "loss": 2.0527,
+      "step": 1357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9967357478632867e-05,
+      "loss": 2.1767,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9967309056042884e-05,
+      "loss": 2.0919,
+      "step": 1359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9967260597622783e-05,
+      "loss": 1.9589,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9967212103372738e-05,
+      "loss": 2.0268,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.996716357329292e-05,
+      "loss": 1.814,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.996711500738351e-05,
+      "loss": 2.0668,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9967066405644676e-05,
+      "loss": 1.9724,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9967017768076595e-05,
+      "loss": 1.5952,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.996696909467944e-05,
+      "loss": 1.908,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9966920385453393e-05,
+      "loss": 1.5855,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9966871640398624e-05,
+      "loss": 1.6964,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.9966822859515306e-05,
+      "loss": 1.9096,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.996677404280362e-05,
+      "loss": 1.9488,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9966725190263737e-05,
+      "loss": 1.788,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9966676301895835e-05,
+      "loss": 2.1291,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9966627377700086e-05,
+      "loss": 2.0096,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.996657841767667e-05,
+      "loss": 1.9399,
+      "step": 1374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.996652942182576e-05,
+      "loss": 1.8639,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9966480390147534e-05,
+      "loss": 1.4761,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.996643132264217e-05,
+      "loss": 1.8941,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9966382219309844e-05,
+      "loss": 2.2227,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.996633308015073e-05,
+      "loss": 1.9034,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9966283905165e-05,
+      "loss": 2.0741,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.996623469435284e-05,
+      "loss": 2.0676,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9966185447714427e-05,
+      "loss": 1.7979,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9966136165249928e-05,
+      "loss": 1.742,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5,
+      "learning_rate": 1.996608684695953e-05,
+      "loss": 2.3651,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9966037492843406e-05,
+      "loss": 1.8799,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9965988102901735e-05,
+      "loss": 1.95,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.996593867713469e-05,
+      "loss": 2.1197,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9965889215542456e-05,
+      "loss": 2.1442,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9965839718125206e-05,
+      "loss": 1.685,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9965790184883122e-05,
+      "loss": 1.927,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9965740615816375e-05,
+      "loss": 1.7325,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.996569101092515e-05,
+      "loss": 1.818,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.996564137020962e-05,
+      "loss": 2.2457,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9965591693669968e-05,
+      "loss": 1.8514,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.996554198130637e-05,
+      "loss": 1.8781,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9965492233119007e-05,
+      "loss": 1.7713,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.996544244910805e-05,
+      "loss": 1.7221,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.996539262927369e-05,
+      "loss": 1.9673,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.99653427736161e-05,
+      "loss": 1.9626,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9965292882135458e-05,
+      "loss": 1.8395,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.996524295483195e-05,
+      "loss": 1.6382,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9965192991705746e-05,
+      "loss": 2.0348,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9965142992757025e-05,
+      "loss": 2.1423,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9965092957985977e-05,
+      "loss": 2.0571,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9965042887392777e-05,
+      "loss": 1.7651,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9964992780977602e-05,
+      "loss": 1.8256,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9964942638740638e-05,
+      "loss": 1.8588,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.996489246068206e-05,
+      "loss": 1.9071,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9964842246802047e-05,
+      "loss": 1.6352,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9964791997100786e-05,
+      "loss": 1.8306,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.9964741711578455e-05,
+      "loss": 1.3329,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9964691390235232e-05,
+      "loss": 2.1157,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.99646410330713e-05,
+      "loss": 1.9089,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.996459064008684e-05,
+      "loss": 1.6441,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9964540211282036e-05,
+      "loss": 2.0162,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9964489746657065e-05,
+      "loss": 1.8998,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.996443924621211e-05,
+      "loss": 1.9461,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.996438870994735e-05,
+      "loss": 2.0104,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9964338137862973e-05,
+      "loss": 1.8872,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9964287529959156e-05,
+      "loss": 1.7216,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.996423688623608e-05,
+      "loss": 1.8763,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.996418620669393e-05,
+      "loss": 2.0724,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.996413549133289e-05,
+      "loss": 1.9629,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9964084740153134e-05,
+      "loss": 1.7023,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9964033953154853e-05,
+      "loss": 1.8994,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9963983130338226e-05,
+      "loss": 1.8253,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9963932271703434e-05,
+      "loss": 1.9177,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9963881377250665e-05,
+      "loss": 1.6541,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.99638304469801e-05,
+      "loss": 2.0336,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9963779480891917e-05,
+      "loss": 1.6723,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9963728478986303e-05,
+      "loss": 1.7936,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9963677441263444e-05,
+      "loss": 2.116,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.996362636772352e-05,
+      "loss": 1.7866,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9963575258366714e-05,
+      "loss": 2.2113,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9963524113193213e-05,
+      "loss": 1.7792,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9963472932203197e-05,
+      "loss": 1.8147,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9963421715396855e-05,
+      "loss": 1.9522,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9963370462774368e-05,
+      "loss": 2.0402,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.996331917433592e-05,
+      "loss": 2.0022,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9963267850081692e-05,
+      "loss": 2.0414,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9963216490011876e-05,
+      "loss": 1.8145,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.9963165094126655e-05,
+      "loss": 3.7585,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9963113662426207e-05,
+      "loss": 2.1259,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9963062194910722e-05,
+      "loss": 1.995,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9963010691580388e-05,
+      "loss": 1.7377,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9962959152435386e-05,
+      "loss": 1.9511,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9962907577475904e-05,
+      "loss": 2.0028,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.996285596670212e-05,
+      "loss": 1.8651,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9962804320114228e-05,
+      "loss": 1.7083,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9962752637712407e-05,
+      "loss": 2.0832,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9962700919496854e-05,
+      "loss": 1.7593,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9962649165467742e-05,
+      "loss": 1.785,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.996259737562526e-05,
+      "loss": 2.0861,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.99625455499696e-05,
+      "loss": 2.3194,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9962493688500948e-05,
+      "loss": 2.0738,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9962441791219482e-05,
+      "loss": 2.0086,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9962389858125396e-05,
+      "loss": 1.9326,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9962337889218876e-05,
+      "loss": 1.6186,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9962285884500104e-05,
+      "loss": 1.8581,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9962233843969273e-05,
+      "loss": 1.856,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9962181767626564e-05,
+      "loss": 1.7949,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.996212965547217e-05,
+      "loss": 1.7345,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9962077507506276e-05,
+      "loss": 1.7654,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.996202532372907e-05,
+      "loss": 1.7737,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9961973104140738e-05,
+      "loss": 1.9308,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9961920848741468e-05,
+      "loss": 1.6401,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9961868557531447e-05,
+      "loss": 1.9364,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9961816230510865e-05,
+      "loss": 1.6621,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9961763867679912e-05,
+      "loss": 1.9653,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9961711469038774e-05,
+      "loss": 1.8051,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.996165903458764e-05,
+      "loss": 1.4902,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.996160656432669e-05,
+      "loss": 1.8583,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9961554058256127e-05,
+      "loss": 1.604,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9961501516376135e-05,
+      "loss": 1.7364,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9961448938686896e-05,
+      "loss": 1.7097,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9961396325188604e-05,
+      "loss": 1.6404,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9961343675881452e-05,
+      "loss": 1.7876,
+      "step": 1477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.996129099076562e-05,
+      "loss": 2.0447,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9961238269841304e-05,
+      "loss": 1.8626,
+      "step": 1479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.9961185513108697e-05,
+      "loss": 2.8561,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.996113272056798e-05,
+      "loss": 1.8301,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9961079892219346e-05,
+      "loss": 2.9113,
+      "step": 1482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9961027028062986e-05,
+      "loss": 1.8666,
+      "step": 1483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.996097412809909e-05,
+      "loss": 1.7351,
+      "step": 1484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9960921192327848e-05,
+      "loss": 2.3124,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.996086822074945e-05,
+      "loss": 2.0384,
+      "step": 1486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9960815213364087e-05,
+      "loss": 1.8098,
+      "step": 1487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9960762170171948e-05,
+      "loss": 1.6383,
+      "step": 1488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9960709091173225e-05,
+      "loss": 1.5373,
+      "step": 1489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9960655976368106e-05,
+      "loss": 1.6346,
+      "step": 1490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.996060282575679e-05,
+      "loss": 1.6814,
+      "step": 1491
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9960549639339456e-05,
+      "loss": 1.8714,
+      "step": 1492
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9960496417116305e-05,
+      "loss": 2.3213,
+      "step": 1493
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9960443159087526e-05,
+      "loss": 1.9201,
+      "step": 1494
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.996038986525331e-05,
+      "loss": 2.0421,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9960336535613845e-05,
+      "loss": 2.0539,
+      "step": 1496
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9960283170169325e-05,
+      "loss": 2.1344,
+      "step": 1497
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9960229768919947e-05,
+      "loss": 1.9518,
+      "step": 1498
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9960176331865896e-05,
+      "loss": 1.7648,
+      "step": 1499
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9960122859007366e-05,
+      "loss": 1.6544,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9960069350344547e-05,
+      "loss": 1.6032,
+      "step": 1501
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.996001580587764e-05,
+      "loss": 2.1915,
+      "step": 1502
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9959962225606827e-05,
+      "loss": 1.8357,
+      "step": 1503
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.995990860953231e-05,
+      "loss": 2.0318,
+      "step": 1504
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9959854957654274e-05,
+      "loss": 1.9113,
+      "step": 1505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9959801269972915e-05,
+      "loss": 1.9818,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9959747546488425e-05,
+      "loss": 1.8672,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9959693787201e-05,
+      "loss": 2.1952,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.995963999211083e-05,
+      "loss": 1.9482,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9959586161218108e-05,
+      "loss": 1.7818,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9959532294523034e-05,
+      "loss": 1.9174,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.9959478392025792e-05,
+      "loss": 2.1991,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9959424453726584e-05,
+      "loss": 2.0997,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9959370479625598e-05,
+      "loss": 1.778,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.995931646972303e-05,
+      "loss": 2.3765,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.995926242401908e-05,
+      "loss": 2.395,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9959208342513932e-05,
+      "loss": 1.5325,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9959154225207786e-05,
+      "loss": 1.9221,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9959100072100837e-05,
+      "loss": 2.1953,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9959045883193275e-05,
+      "loss": 1.8805,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9958991658485302e-05,
+      "loss": 1.8505,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9958937397977106e-05,
+      "loss": 1.8233,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.995888310166889e-05,
+      "loss": 2.1648,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.995882876956084e-05,
+      "loss": 2.2423,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.995877440165316e-05,
+      "loss": 2.0339,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9958719997946037e-05,
+      "loss": 2.2179,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.995866555843967e-05,
+      "loss": 2.0749,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.995861108313426e-05,
+      "loss": 1.6381,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9958556572029995e-05,
+      "loss": 1.9828,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9958502025127074e-05,
+      "loss": 1.8305,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9958447442425695e-05,
+      "loss": 1.663,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9958392823926054e-05,
+      "loss": 1.8768,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9958338169628345e-05,
+      "loss": 1.8588,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9958283479532767e-05,
+      "loss": 2.1247,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.995822875363951e-05,
+      "loss": 1.8606,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.995817399194878e-05,
+      "loss": 1.8174,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9958119194460767e-05,
+      "loss": 1.8969,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.995806436117567e-05,
+      "loss": 1.7847,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.995800949209369e-05,
+      "loss": 1.9731,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9957954587215015e-05,
+      "loss": 1.6368,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9957899646539857e-05,
+      "loss": 2.0166,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9957844670068396e-05,
+      "loss": 1.9481,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9957789657800844e-05,
+      "loss": 1.4234,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9957734609737392e-05,
+      "loss": 1.745,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9957679525878236e-05,
+      "loss": 1.9446,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.995762440622358e-05,
+      "loss": 2.1908,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.995756925077362e-05,
+      "loss": 1.947,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9957514059528552e-05,
+      "loss": 1.8207,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9957458832488573e-05,
+      "loss": 1.8539,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.995740356965389e-05,
+      "loss": 2.1357,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9957348271024692e-05,
+      "loss": 1.8597,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9957292936601183e-05,
+      "loss": 1.9934,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.995723756638356e-05,
+      "loss": 2.0546,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9957182160372024e-05,
+      "loss": 1.7273,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9957126718566774e-05,
+      "loss": 1.7789,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9957071240968005e-05,
+      "loss": 1.9156,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.995701572757592e-05,
+      "loss": 1.7614,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9956960178390722e-05,
+      "loss": 2.0026,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9956904593412604e-05,
+      "loss": 2.1706,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.995684897264177e-05,
+      "loss": 1.6526,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9956793316078417e-05,
+      "loss": 1.8351,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.995673762372275e-05,
+      "loss": 1.8983,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.995668189557496e-05,
+      "loss": 1.9327,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9956626131635258e-05,
+      "loss": 2.0415,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.995657033190384e-05,
+      "loss": 1.7518,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9956514496380907e-05,
+      "loss": 1.8904,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9956458625066655e-05,
+      "loss": 1.9983,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.995640271796129e-05,
+      "loss": 1.5789,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9956346775065013e-05,
+      "loss": 2.194,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9956290796378022e-05,
+      "loss": 1.9079,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9956234781900522e-05,
+      "loss": 1.8833,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9956178731632715e-05,
+      "loss": 2.0838,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9956122645574798e-05,
+      "loss": 2.1626,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.995606652372697e-05,
+      "loss": 1.9255,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9956010366089443e-05,
+      "loss": 1.8572,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9955954172662412e-05,
+      "loss": 2.1994,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9955897943446078e-05,
+      "loss": 1.9079,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.995584167844065e-05,
+      "loss": 1.9467,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.995578537764632e-05,
+      "loss": 1.8494,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9955729041063302e-05,
+      "loss": 1.6316,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.995567266869179e-05,
+      "loss": 1.9155,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9955616260531986e-05,
+      "loss": 1.9186,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.99555598165841e-05,
+      "loss": 1.8613,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.995550333684833e-05,
+      "loss": 1.7295,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.995544682132488e-05,
+      "loss": 1.8196,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9955390270013952e-05,
+      "loss": 2.0829,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9955333682915752e-05,
+      "loss": 2.0192,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.995527706003048e-05,
+      "loss": 1.639,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9955220401358342e-05,
+      "loss": 1.9367,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9955163706899542e-05,
+      "loss": 2.0975,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9955106976654282e-05,
+      "loss": 1.95,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9955050210622767e-05,
+      "loss": 2.174,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.99549934088052e-05,
+      "loss": 2.0246,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9954936571201787e-05,
+      "loss": 1.9797,
+      "step": 1594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.995487969781273e-05,
+      "loss": 1.9056,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9954822788638236e-05,
+      "loss": 1.9283,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.995476584367851e-05,
+      "loss": 1.6423,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.995470886293375e-05,
+      "loss": 2.5506,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.995465184640417e-05,
+      "loss": 1.7169,
+      "step": 1599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.995459479408997e-05,
+      "loss": 1.9234,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.995453770599135e-05,
+      "loss": 1.8645,
+      "step": 1601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.995448058210853e-05,
+      "loss": 1.9867,
+      "step": 1602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9954423422441703e-05,
+      "loss": 1.8312,
+      "step": 1603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9954366226991078e-05,
+      "loss": 1.8353,
+      "step": 1604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.995430899575686e-05,
+      "loss": 1.6714,
+      "step": 1605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9954251728739256e-05,
+      "loss": 2.054,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9954194425938473e-05,
+      "loss": 1.953,
+      "step": 1607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.995413708735471e-05,
+      "loss": 2.065,
+      "step": 1608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9954079712988183e-05,
+      "loss": 1.8379,
+      "step": 1609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9954022302839093e-05,
+      "loss": 2.2597,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9953964856907644e-05,
+      "loss": 1.7922,
+      "step": 1611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9953907375194048e-05,
+      "loss": 2.1803,
+      "step": 1612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9953849857698508e-05,
+      "loss": 1.9665,
+      "step": 1613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9953792304421232e-05,
+      "loss": 2.0223,
+      "step": 1614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9953734715362426e-05,
+      "loss": 1.7874,
+      "step": 1615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9953677090522298e-05,
+      "loss": 1.6222,
+      "step": 1616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9953619429901056e-05,
+      "loss": 2.1188,
+      "step": 1617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.995356173349891e-05,
+      "loss": 1.7765,
+      "step": 1618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9953504001316054e-05,
+      "loss": 1.9814,
+      "step": 1619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.995344623335271e-05,
+      "loss": 1.5748,
+      "step": 1620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.995338842960908e-05,
+      "loss": 1.7756,
+      "step": 1621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9953330590085374e-05,
+      "loss": 1.8484,
+      "step": 1622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9953272714781798e-05,
+      "loss": 2.1288,
+      "step": 1623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.995321480369856e-05,
+      "loss": 1.6358,
+      "step": 1624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9953156856835867e-05,
+      "loss": 2.0822,
+      "step": 1625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.995309887419393e-05,
+      "loss": 1.7454,
+      "step": 1626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.995304085577296e-05,
+      "loss": 1.9211,
+      "step": 1627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9952982801573158e-05,
+      "loss": 1.8088,
+      "step": 1628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.995292471159474e-05,
+      "loss": 1.6791,
+      "step": 1629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9952866585837907e-05,
+      "loss": 1.7198,
+      "step": 1630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9952808424302874e-05,
+      "loss": 2.0221,
+      "step": 1631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.995275022698985e-05,
+      "loss": 1.8699,
+      "step": 1632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.995269199389904e-05,
+      "loss": 1.7121,
+      "step": 1633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9952633725030658e-05,
+      "loss": 2.0204,
+      "step": 1634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9952575420384912e-05,
+      "loss": 1.9314,
+      "step": 1635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.995251707996201e-05,
+      "loss": 1.6837,
+      "step": 1636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9952458703762163e-05,
+      "loss": 2.0784,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.995240029178558e-05,
+      "loss": 1.977,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9952341844032473e-05,
+      "loss": 2.0574,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.995228336050305e-05,
+      "loss": 1.9149,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.995222484119752e-05,
+      "loss": 1.777,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9952166286116097e-05,
+      "loss": 1.9532,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9952107695258993e-05,
+      "loss": 2.0636,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.995204906862641e-05,
+      "loss": 1.8599,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9951990406218567e-05,
+      "loss": 1.8746,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9951931708035673e-05,
+      "loss": 1.7593,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9951872974077936e-05,
+      "loss": 2.0731,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9951814204345573e-05,
+      "loss": 1.7779,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9951755398838787e-05,
+      "loss": 1.8535,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9951696557557794e-05,
+      "loss": 1.7884,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9951637680502806e-05,
+      "loss": 1.7894,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9951578767674034e-05,
+      "loss": 1.7134,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9951519819071694e-05,
+      "loss": 2.0155,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.995146083469599e-05,
+      "loss": 2.4436,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.995140181454714e-05,
+      "loss": 1.9499,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.995134275862535e-05,
+      "loss": 1.8983,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.995128366693084e-05,
+      "loss": 2.0947,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9951224539463814e-05,
+      "loss": 1.9932,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9951165376224494e-05,
+      "loss": 1.8942,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9951106177213085e-05,
+      "loss": 2.0016,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9951046942429804e-05,
+      "loss": 2.0015,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9950987671874862e-05,
+      "loss": 2.2412,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9950928365548475e-05,
+      "loss": 2.2528,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.995086902345085e-05,
+      "loss": 1.6711,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9950809645582207e-05,
+      "loss": 2.1182,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9950750231942756e-05,
+      "loss": 2.0246,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9950690782532708e-05,
+      "loss": 1.9947,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9950631297352285e-05,
+      "loss": 2.1493,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.995057177640169e-05,
+      "loss": 2.1118,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9950512219681147e-05,
+      "loss": 2.035,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9950452627190865e-05,
+      "loss": 1.8869,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.995039299893106e-05,
+      "loss": 1.9579,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9950333334901945e-05,
+      "loss": 1.7741,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9950273635103736e-05,
+      "loss": 2.0544,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.995021389953664e-05,
+      "loss": 1.7588,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.9950154128200885e-05,
+      "loss": 1.9356,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9950094321096676e-05,
+      "loss": 1.9524,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.995003447822423e-05,
+      "loss": 1.8191,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9949974599583767e-05,
+      "loss": 1.8535,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9949914685175497e-05,
+      "loss": 1.7578,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9949854734999637e-05,
+      "loss": 1.8122,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9949794749056403e-05,
+      "loss": 1.8212,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9949734727346005e-05,
+      "loss": 1.761,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.994967466986867e-05,
+      "loss": 1.4237,
+      "step": 1684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9949614576624602e-05,
+      "loss": 1.7156,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9949554447614026e-05,
+      "loss": 1.8063,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9949494282837155e-05,
+      "loss": 2.8986,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9949434082294203e-05,
+      "loss": 1.6797,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9949373845985393e-05,
+      "loss": 1.7983,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9949313573910932e-05,
+      "loss": 2.3088,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.994925326607104e-05,
+      "loss": 1.7409,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.994919292246594e-05,
+      "loss": 1.9391,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9949132543095844e-05,
+      "loss": 2.0285,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9949072127960968e-05,
+      "loss": 1.8878,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.994901167706153e-05,
+      "loss": 2.0686,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.994895119039775e-05,
+      "loss": 1.81,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9948890667969844e-05,
+      "loss": 1.6061,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9948830109778025e-05,
+      "loss": 1.9287,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9948769515822516e-05,
+      "loss": 1.6614,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9948708886103537e-05,
+      "loss": 1.9454,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9948648220621297e-05,
+      "loss": 1.9228,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9948587519376024e-05,
+      "loss": 1.7134,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.994852678236793e-05,
+      "loss": 2.148,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9948466009597233e-05,
+      "loss": 1.8487,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9948405201064155e-05,
+      "loss": 1.7813,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9948344356768912e-05,
+      "loss": 2.325,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9948283476711725e-05,
+      "loss": 2.4778,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.994822256089281e-05,
+      "loss": 1.7813,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.994816160931239e-05,
+      "loss": 2.0342,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9948100621970682e-05,
+      "loss": 1.6595,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9948039598867902e-05,
+      "loss": 1.9664,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9947978540004276e-05,
+      "loss": 2.0324,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.994791744538002e-05,
+      "loss": 1.9609,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.994785631499535e-05,
+      "loss": 1.8631,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9947795148850488e-05,
+      "loss": 1.5728,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.994773394694566e-05,
+      "loss": 1.8475,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9947672709281076e-05,
+      "loss": 2.7842,
+      "step": 1717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9947611435856963e-05,
+      "loss": 1.8117,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9947550126673542e-05,
+      "loss": 1.9302,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.994748878173103e-05,
+      "loss": 2.1566,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9947427401029645e-05,
+      "loss": 1.8235,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.9947365984569613e-05,
+      "loss": 1.9035,
+      "step": 1722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9947304532351153e-05,
+      "loss": 2.0267,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9947243044374486e-05,
+      "loss": 1.9714,
+      "step": 1724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9947181520639832e-05,
+      "loss": 1.7288,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.994711996114741e-05,
+      "loss": 1.8869,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.994705836589745e-05,
+      "loss": 2.1164,
+      "step": 1727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9946996734890166e-05,
+      "loss": 1.9168,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.994693506812578e-05,
+      "loss": 1.8006,
+      "step": 1729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9946873365604512e-05,
+      "loss": 1.8169,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.994681162732659e-05,
+      "loss": 1.8883,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9946749853292233e-05,
+      "loss": 1.994,
+      "step": 1732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.994668804350166e-05,
+      "loss": 1.9387,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.99466261979551e-05,
+      "loss": 1.9731,
+      "step": 1734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9946564316652765e-05,
+      "loss": 2.0765,
+      "step": 1735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.994650239959489e-05,
+      "loss": 1.8587,
+      "step": 1736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.994644044678169e-05,
+      "loss": 1.6707,
+      "step": 1737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9946378458213385e-05,
+      "loss": 1.9628,
+      "step": 1738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9946316433890204e-05,
+      "loss": 2.177,
+      "step": 1739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9946254373812367e-05,
+      "loss": 1.9622,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.99461922779801e-05,
+      "loss": 1.8568,
+      "step": 1741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.994613014639362e-05,
+      "loss": 1.5189,
+      "step": 1742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.9946067979053158e-05,
+      "loss": 1.767,
+      "step": 1743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9946005775958933e-05,
+      "loss": 1.8435,
+      "step": 1744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9945943537111167e-05,
+      "loss": 1.9993,
+      "step": 1745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9945881262510086e-05,
+      "loss": 1.9125,
+      "step": 1746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9945818952155917e-05,
+      "loss": 1.9407,
+      "step": 1747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.994575660604888e-05,
+      "loss": 1.7779,
+      "step": 1748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.99456942241892e-05,
+      "loss": 2.1474,
+      "step": 1749
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.99456318065771e-05,
+      "loss": 2.0415,
+      "step": 1750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.994556935321281e-05,
+      "loss": 1.6954,
+      "step": 1751
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9945506864096545e-05,
+      "loss": 1.824,
+      "step": 1752
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9945444339228538e-05,
+      "loss": 1.7749,
+      "step": 1753
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.994538177860901e-05,
+      "loss": 1.7875,
+      "step": 1754
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.994531918223819e-05,
+      "loss": 1.8434,
+      "step": 1755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.994525655011629e-05,
+      "loss": 1.6063,
+      "step": 1756
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9945193882243552e-05,
+      "loss": 2.1206,
+      "step": 1757
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9945131178620195e-05,
+      "loss": 1.6964,
+      "step": 1758
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.994506843924644e-05,
+      "loss": 1.5431,
+      "step": 1759
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9945005664122515e-05,
+      "loss": 1.9639,
+      "step": 1760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.994494285324865e-05,
+      "loss": 1.7267,
+      "step": 1761
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9944880006625063e-05,
+      "loss": 2.136,
+      "step": 1762
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.994481712425199e-05,
+      "loss": 1.9058,
+      "step": 1763
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9944754206129645e-05,
+      "loss": 1.9864,
+      "step": 1764
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9944691252258264e-05,
+      "loss": 2.0093,
+      "step": 1765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.994462826263807e-05,
+      "loss": 1.9607,
+      "step": 1766
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.994456523726929e-05,
+      "loss": 2.1533,
+      "step": 1767
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9944502176152147e-05,
+      "loss": 1.9347,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9944439079286873e-05,
+      "loss": 1.9956,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.994437594667369e-05,
+      "loss": 1.9787,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.994431277831283e-05,
+      "loss": 2.1028,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9944249574204514e-05,
+      "loss": 2.0345,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9944186334348976e-05,
+      "loss": 2.135,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9944123058746438e-05,
+      "loss": 1.8204,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.994405974739713e-05,
+      "loss": 2.0576,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.994399640030128e-05,
+      "loss": 1.9558,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9943933017459112e-05,
+      "loss": 2.7491,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9943869598870856e-05,
+      "loss": 2.0678,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9943806144536743e-05,
+      "loss": 1.6869,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9943742654457e-05,
+      "loss": 1.8521,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.994367912863185e-05,
+      "loss": 1.8823,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9943615567061525e-05,
+      "loss": 2.0091,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.994355196974626e-05,
+      "loss": 1.7841,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.994348833668627e-05,
+      "loss": 2.1135,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.994342466788179e-05,
+      "loss": 1.7626,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9943360963333055e-05,
+      "loss": 1.7412,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9943297223040286e-05,
+      "loss": 1.6999,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9943233447003717e-05,
+      "loss": 1.6464,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9943169635223574e-05,
+      "loss": 1.7468,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.994310578770009e-05,
+      "loss": 1.8679,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.9943041904433488e-05,
+      "loss": 1.7777,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9942977985424008e-05,
+      "loss": 1.7852,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.994291403067187e-05,
+      "loss": 1.9899,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9942850040177305e-05,
+      "loss": 1.9469,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9942786013940546e-05,
+      "loss": 2.0413,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9942721951961826e-05,
+      "loss": 1.8419,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.994265785424137e-05,
+      "loss": 1.8724,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9942593720779412e-05,
+      "loss": 1.7289,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.994252955157618e-05,
+      "loss": 1.9956,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9942465346631905e-05,
+      "loss": 2.005,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9942401105946817e-05,
+      "loss": 1.9108,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9942336829521153e-05,
+      "loss": 2.2174,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9942272517355134e-05,
+      "loss": 1.7268,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9942208169449e-05,
+      "loss": 1.7848,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.994214378580298e-05,
+      "loss": 1.7374,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.99420793664173e-05,
+      "loss": 2.2133,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.99420149112922e-05,
+      "loss": 1.8717,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9941950420427903e-05,
+      "loss": 1.8066,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9941885893824652e-05,
+      "loss": 2.1462,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9941821331482668e-05,
+      "loss": 1.8002,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9941756733402186e-05,
+      "loss": 2.014,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.994169209958344e-05,
+      "loss": 2.2718,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9941627430026667e-05,
+      "loss": 2.2545,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.994156272473209e-05,
+      "loss": 1.7309,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.994149798369995e-05,
+      "loss": 1.6809,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9941433206930473e-05,
+      "loss": 1.8977,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.994136839442389e-05,
+      "loss": 2.1346,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9941303546180448e-05,
+      "loss": 2.0213,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9941238662200365e-05,
+      "loss": 1.6144,
+      "step": 1819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.994117374248388e-05,
+      "loss": 2.0966,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.994110878703123e-05,
+      "loss": 1.755,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9941043795842642e-05,
+      "loss": 1.8249,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9940978768918353e-05,
+      "loss": 1.7645,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.99409137062586e-05,
+      "loss": 1.9436,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.994084860786361e-05,
+      "loss": 1.5769,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.994078347373362e-05,
+      "loss": 2.0286,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9940718303868865e-05,
+      "loss": 1.9441,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9940653098269576e-05,
+      "loss": 2.06,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.994058785693599e-05,
+      "loss": 2.2802,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9940522579868347e-05,
+      "loss": 1.9657,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.994045726706687e-05,
+      "loss": 2.2898,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9940391918531806e-05,
+      "loss": 1.9086,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.994032653426338e-05,
+      "loss": 2.0949,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.994026111426183e-05,
+      "loss": 1.8077,
+      "step": 1834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9940195658527393e-05,
+      "loss": 2.038,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9940130167060306e-05,
+      "loss": 1.7894,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.99400646398608e-05,
+      "loss": 1.8783,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9939999076929114e-05,
+      "loss": 1.8309,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9939933478265477e-05,
+      "loss": 2.2327,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9939867843870135e-05,
+      "loss": 2.0154,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9939802173743317e-05,
+      "loss": 1.8746,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.993973646788526e-05,
+      "loss": 2.1853,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9939670726296203e-05,
+      "loss": 1.5066,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.993960494897638e-05,
+      "loss": 1.739,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9939539135926028e-05,
+      "loss": 1.8328,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9939473287145383e-05,
+      "loss": 2.0479,
+      "step": 1846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9939407402634682e-05,
+      "loss": 2.0784,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.993934148239416e-05,
+      "loss": 1.901,
+      "step": 1848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9939275526424057e-05,
+      "loss": 1.772,
+      "step": 1849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.993920953472461e-05,
+      "loss": 1.7866,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9939143507296056e-05,
+      "loss": 1.8979,
+      "step": 1851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9939077444138633e-05,
+      "loss": 2.0382,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.993901134525257e-05,
+      "loss": 2.1273,
+      "step": 1853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.993894521063812e-05,
+      "loss": 2.0113,
+      "step": 1854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9938879040295508e-05,
+      "loss": 1.7257,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9938812834224978e-05,
+      "loss": 1.6943,
+      "step": 1856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9938746592426765e-05,
+      "loss": 1.9505,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.993868031490111e-05,
+      "loss": 2.535,
+      "step": 1858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.993861400164825e-05,
+      "loss": 1.8689,
+      "step": 1859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.993854765266842e-05,
+      "loss": 1.8546,
+      "step": 1860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9938481267961865e-05,
+      "loss": 1.7594,
+      "step": 1861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9938414847528818e-05,
+      "loss": 1.8903,
+      "step": 1862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9938348391369523e-05,
+      "loss": 2.1302,
+      "step": 1863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9938281899484212e-05,
+      "loss": 2.0401,
+      "step": 1864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.993821537187313e-05,
+      "loss": 2.0161,
+      "step": 1865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9938148808536512e-05,
+      "loss": 1.5977,
+      "step": 1866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.99380822094746e-05,
+      "loss": 1.9024,
+      "step": 1867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.75,
+      "learning_rate": 1.993801557468764e-05,
+      "loss": 2.3967,
+      "step": 1868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9937948904175856e-05,
+      "loss": 1.866,
+      "step": 1869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9937882197939496e-05,
+      "loss": 2.0318,
+      "step": 1870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9937815455978803e-05,
+      "loss": 1.7958,
+      "step": 1871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.993774867829401e-05,
+      "loss": 1.903,
+      "step": 1872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9937681864885365e-05,
+      "loss": 1.5524,
+      "step": 1873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.99376150157531e-05,
+      "loss": 1.8846,
+      "step": 1874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9937548130897464e-05,
+      "loss": 2.3418,
+      "step": 1875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.993748121031869e-05,
+      "loss": 2.0098,
+      "step": 1876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.993741425401702e-05,
+      "loss": 1.7418,
+      "step": 1877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9937347261992696e-05,
+      "loss": 1.9811,
+      "step": 1878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.993728023424596e-05,
+      "loss": 2.1345,
+      "step": 1879
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.993721317077705e-05,
+      "loss": 1.8711,
+      "step": 1880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9937146071586212e-05,
+      "loss": 2.1648,
+      "step": 1881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.993707893667368e-05,
+      "loss": 1.8868,
+      "step": 1882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9937011766039703e-05,
+      "loss": 2.2301,
+      "step": 1883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9936944559684517e-05,
+      "loss": 1.9679,
+      "step": 1884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9936877317608366e-05,
+      "loss": 2.2069,
+      "step": 1885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.993681003981149e-05,
+      "loss": 1.9775,
+      "step": 1886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.993674272629413e-05,
+      "loss": 1.9779,
+      "step": 1887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9936675377056534e-05,
+      "loss": 1.9201,
+      "step": 1888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9936607992098938e-05,
+      "loss": 1.9534,
+      "step": 1889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9936540571421586e-05,
+      "loss": 1.9999,
+      "step": 1890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.993647311502472e-05,
+      "loss": 1.8146,
+      "step": 1891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9936405622908583e-05,
+      "loss": 1.6374,
+      "step": 1892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.993633809507342e-05,
+      "loss": 1.8318,
+      "step": 1893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9936270531519472e-05,
+      "loss": 1.6887,
+      "step": 1894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9936202932246975e-05,
+      "loss": 1.5921,
+      "step": 1895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9936135297256183e-05,
+      "loss": 1.8175,
+      "step": 1896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9936067626547334e-05,
+      "loss": 1.8454,
+      "step": 1897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9935999920120673e-05,
+      "loss": 1.8535,
+      "step": 1898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9935932177976443e-05,
+      "loss": 1.7967,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9935864400114883e-05,
+      "loss": 2.0238,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9935796586536244e-05,
+      "loss": 2.0337,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.993572873724076e-05,
+      "loss": 1.7414,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9935660852228685e-05,
+      "loss": 1.9123,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.993559293150026e-05,
+      "loss": 2.2868,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9935524975055728e-05,
+      "loss": 2.0514,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.993545698289533e-05,
+      "loss": 2.0849,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9935388955019314e-05,
+      "loss": 1.8854,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.993532089142793e-05,
+      "loss": 1.6915,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.993525279212141e-05,
+      "loss": 1.9718,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.993518465710001e-05,
+      "loss": 1.7518,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.993511648636397e-05,
+      "loss": 1.8219,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9935048279913536e-05,
+      "loss": 2.1577,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9934980037748955e-05,
+      "loss": 2.0004,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9934911759870466e-05,
+      "loss": 1.8283,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.993484344627832e-05,
+      "loss": 1.7085,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9934775096972763e-05,
+      "loss": 1.8353,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.993470671195404e-05,
+      "loss": 1.9147,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9934638291222393e-05,
+      "loss": 1.7438,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.993456983477807e-05,
+      "loss": 1.8247,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.993450134262132e-05,
+      "loss": 1.8233,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.993443281475239e-05,
+      "loss": 1.8095,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.993436425117152e-05,
+      "loss": 1.7352,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9934295651878958e-05,
+      "loss": 2.1603,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9934227016874956e-05,
+      "loss": 1.9549,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9934158346159755e-05,
+      "loss": 1.6616,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9934089639733606e-05,
+      "loss": 1.5132,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9934020897596752e-05,
+      "loss": 2.053,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9933952119749443e-05,
+      "loss": 1.7771,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9933883306191927e-05,
+      "loss": 1.9768,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9933814456924446e-05,
+      "loss": 1.7292,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9933745571947256e-05,
+      "loss": 1.9643,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9933676651260597e-05,
+      "loss": 2.0144,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.993360769486472e-05,
+      "loss": 2.0947,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9933538702759874e-05,
+      "loss": 1.9191,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9933469674946305e-05,
+      "loss": 1.9476,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9933400611424263e-05,
+      "loss": 2.257,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9933331512193992e-05,
+      "loss": 1.9715,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9933262377255746e-05,
+      "loss": 2.0034,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.993319320660977e-05,
+      "loss": 1.9561,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9933124000256315e-05,
+      "loss": 1.9238,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9933054758195625e-05,
+      "loss": 1.8931,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9932985480427953e-05,
+      "loss": 1.9152,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.993291616695355e-05,
+      "loss": 1.8807,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.993284681777266e-05,
+      "loss": 1.6028,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.9932777432885536e-05,
+      "loss": 2.3028,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9932708012292425e-05,
+      "loss": 1.7102,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.993263855599358e-05,
+      "loss": 2.0664,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.993256906398924e-05,
+      "loss": 2.2209,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9932499536279676e-05,
+      "loss": 1.8558,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9932429972865118e-05,
+      "loss": 1.8725,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9932360373745824e-05,
+      "loss": 1.9051,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.993229073892204e-05,
+      "loss": 1.8743,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9932221068394027e-05,
+      "loss": 1.8153,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9932151362162023e-05,
+      "loss": 1.9488,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9932081620226286e-05,
+      "loss": 1.7383,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9932011842587063e-05,
+      "loss": 2.2357,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9931942029244605e-05,
+      "loss": 2.0428,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.9931872180199166e-05,
+      "loss": 1.7014,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9931802295450997e-05,
+      "loss": 2.1478,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9931732375000346e-05,
+      "loss": 1.9061,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9931662418847466e-05,
+      "loss": 1.7277,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9931592426992608e-05,
+      "loss": 2.1677,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9931522399436026e-05,
+      "loss": 1.8075,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9931452336177967e-05,
+      "loss": 1.5884,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9931382237218686e-05,
+      "loss": 2.1303,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9931312102558437e-05,
+      "loss": 1.9471,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 1.9931241932197464e-05,
+      "loss": 2.1176,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.993117172613603e-05,
+      "loss": 2.021,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9931101484374384e-05,
+      "loss": 2.0505,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9931031206912772e-05,
+      "loss": 2.1918,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9930960893751453e-05,
+      "loss": 1.9709,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.993089054489068e-05,
+      "loss": 1.6075,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9930820160330703e-05,
+      "loss": 2.225,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.9930749740071774e-05,
+      "loss": 2.0312,
+      "step": 1974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.993067928411415e-05,
+      "loss": 2.5006,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9930608792458083e-05,
+      "loss": 1.8625,
+      "step": 1976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9930538265103825e-05,
+      "loss": 1.7075,
+      "step": 1977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9930467702051632e-05,
+      "loss": 2.2324,
+      "step": 1978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9930397103301753e-05,
+      "loss": 1.8554,
+      "step": 1979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.993032646885445e-05,
+      "loss": 2.0789,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9930255798709966e-05,
+      "loss": 1.9906,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25,
+      "learning_rate": 1.993018509286856e-05,
+      "loss": 2.0591,
+      "step": 1982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9930114351330495e-05,
+      "loss": 1.6019,
+      "step": 1983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9930043574096013e-05,
+      "loss": 1.9108,
+      "step": 1984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.992997276116537e-05,
+      "loss": 2.0,
+      "step": 1985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9929901912538826e-05,
+      "loss": 1.497,
+      "step": 1986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9929831028216634e-05,
+      "loss": 2.0366,
+      "step": 1987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9929760108199044e-05,
+      "loss": 1.9084,
+      "step": 1988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.992968915248632e-05,
+      "loss": 2.1804,
+      "step": 1989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9929618161078706e-05,
+      "loss": 1.7583,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9929547133976466e-05,
+      "loss": 1.9706,
+      "step": 1991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9929476071179855e-05,
+      "loss": 2.0498,
+      "step": 1992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9929404972689123e-05,
+      "loss": 1.9854,
+      "step": 1993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9929333838504528e-05,
+      "loss": 2.0722,
+      "step": 1994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.992926266862633e-05,
+      "loss": 2.1604,
+      "step": 1995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9929191463054773e-05,
+      "loss": 1.6147,
+      "step": 1996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9929120221790126e-05,
+      "loss": 1.9234,
+      "step": 1997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9929048944832638e-05,
+      "loss": 2.1356,
+      "step": 1998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9928977632182568e-05,
+      "loss": 1.7443,
+      "step": 1999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9928906283840176e-05,
+      "loss": 2.0759,
+      "step": 2000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.992883489980571e-05,
+      "loss": 2.0293,
+      "step": 2001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.992876348007943e-05,
+      "loss": 1.994,
+      "step": 2002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.9928692024661594e-05,
+      "loss": 1.6741,
+      "step": 2003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.992862053355246e-05,
+      "loss": 1.987,
+      "step": 2004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9928549006752277e-05,
+      "loss": 1.9601,
+      "step": 2005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9928477444261315e-05,
+      "loss": 1.9138,
+      "step": 2006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9928405846079823e-05,
+      "loss": 1.9384,
+      "step": 2007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.992833421220806e-05,
+      "loss": 1.654,
+      "step": 2008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.992826254264628e-05,
+      "loss": 1.9995,
+      "step": 2009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.992819083739475e-05,
+      "loss": 1.8309,
+      "step": 2010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9928119096453717e-05,
+      "loss": 1.9572,
+      "step": 2011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9928047319823447e-05,
+      "loss": 1.7028,
+      "step": 2012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9927975507504193e-05,
+      "loss": 2.0905,
+      "step": 2013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9927903659496218e-05,
+      "loss": 2.1061,
+      "step": 2014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9927831775799774e-05,
+      "loss": 1.7688,
+      "step": 2015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.992775985641512e-05,
+      "loss": 1.9744,
+      "step": 2016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.992768790134252e-05,
+      "loss": 1.8424,
+      "step": 2017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9927615910582234e-05,
+      "loss": 2.0013,
+      "step": 2018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9927543884134514e-05,
+      "loss": 1.987,
+      "step": 2019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.992747182199962e-05,
+      "loss": 1.9283,
+      "step": 2020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9927399724177814e-05,
+      "loss": 1.665,
+      "step": 2021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.992732759066935e-05,
+      "loss": 1.9483,
+      "step": 2022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9927255421474496e-05,
+      "loss": 1.7117,
+      "step": 2023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9927183216593505e-05,
+      "loss": 1.7534,
+      "step": 2024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9927110976026638e-05,
+      "loss": 2.0806,
+      "step": 2025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9927038699774153e-05,
+      "loss": 1.6736,
+      "step": 2026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9926966387836313e-05,
+      "loss": 2.1139,
+      "step": 2027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9926894040213375e-05,
+      "loss": 2.0372,
+      "step": 2028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9926821656905605e-05,
+      "loss": 1.8794,
+      "step": 2029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9926749237913253e-05,
+      "loss": 1.9319,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.992667678323659e-05,
+      "loss": 1.8913,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.992660429287587e-05,
+      "loss": 1.8835,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9926531766831353e-05,
+      "loss": 1.7367,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9926459205103303e-05,
+      "loss": 1.9151,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.992638660769198e-05,
+      "loss": 1.7734,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9926313974597645e-05,
+      "loss": 1.9907,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9926241305820556e-05,
+      "loss": 1.9699,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.992616860136098e-05,
+      "loss": 1.9045,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9926095861219176e-05,
+      "loss": 2.0608,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9926023085395404e-05,
+      "loss": 2.0115,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9925950273889928e-05,
+      "loss": 1.8871,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9925877426703004e-05,
+      "loss": 1.9888,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9925804543834897e-05,
+      "loss": 2.3868,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9925731625285877e-05,
+      "loss": 1.9169,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9925658671056194e-05,
+      "loss": 2.0626,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.992558568114612e-05,
+      "loss": 2.121,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.992551265555591e-05,
+      "loss": 1.7854,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9925439594285828e-05,
+      "loss": 1.9153,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9925366497336138e-05,
+      "loss": 2.2721,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9925293364707103e-05,
+      "loss": 1.754,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9925220196398987e-05,
+      "loss": 1.8711,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.992514699241205e-05,
+      "loss": 2.5993,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.992507375274656e-05,
+      "loss": 1.9379,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9925000477402775e-05,
+      "loss": 1.7909,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.992492716638096e-05,
+      "loss": 1.9472,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.992485381968138e-05,
+      "loss": 1.872,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9924780437304298e-05,
+      "loss": 1.925,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9924707019249974e-05,
+      "loss": 1.7394,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.992463356551868e-05,
+      "loss": 1.9219,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9924560076110674e-05,
+      "loss": 2.204,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.992448655102622e-05,
+      "loss": 1.9084,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9924412990265583e-05,
+      "loss": 1.9764,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.992433939382903e-05,
+      "loss": 1.8449,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9924265761716827e-05,
+      "loss": 2.0478,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.992419209392923e-05,
+      "loss": 2.3122,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9924118390466512e-05,
+      "loss": 1.8281,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9924044651328937e-05,
+      "loss": 2.3686,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9923970876516767e-05,
+      "loss": 2.2332,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.992389706603027e-05,
+      "loss": 2.0584,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9923823219869706e-05,
+      "loss": 2.0254,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.992374933803535e-05,
+      "loss": 2.4023,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.992367542052746e-05,
+      "loss": 1.9121,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.99236014673463e-05,
+      "loss": 1.9902,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.992352747849214e-05,
+      "loss": 2.2041,
+      "step": 2074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9923453453965253e-05,
+      "loss": 1.8617,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.992337939376589e-05,
+      "loss": 1.6041,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9923305297894328e-05,
+      "loss": 1.8815,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.992323116635083e-05,
+      "loss": 1.7357,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9923156999135666e-05,
+      "loss": 2.4047,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.992308279624909e-05,
+      "loss": 2.0904,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9923008557691386e-05,
+      "loss": 2.0562,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9922934283462813e-05,
+      "loss": 2.1088,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9922859973563637e-05,
+      "loss": 1.8888,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9922785627994124e-05,
+      "loss": 1.5377,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9922711246754543e-05,
+      "loss": 1.6973,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9922636829845165e-05,
+      "loss": 1.8707,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.992256237726625e-05,
+      "loss": 2.1773,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9922487889018072e-05,
+      "loss": 1.7854,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9922413365100897e-05,
+      "loss": 1.788,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9922338805514995e-05,
+      "loss": 1.7953,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9922264210260627e-05,
+      "loss": 1.8711,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9922189579338068e-05,
+      "loss": 1.8123,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9922114912747585e-05,
+      "loss": 1.9814,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9922040210489442e-05,
+      "loss": 1.9245,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9921965472563914e-05,
+      "loss": 1.9566,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9921890698971265e-05,
+      "loss": 1.9689,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9921815889711767e-05,
+      "loss": 1.7394,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9921741044785686e-05,
+      "loss": 2.1898,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.992166616419329e-05,
+      "loss": 1.8805,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9921591247934856e-05,
+      "loss": 1.9585,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9921516296010645e-05,
+      "loss": 1.9042,
+      "step": 2101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.992144130842093e-05,
+      "loss": 2.0553,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.992136628516598e-05,
+      "loss": 1.9062,
+      "step": 2103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9921291226246062e-05,
+      "loss": 1.6707,
+      "step": 2104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.992121613166145e-05,
+      "loss": 2.124,
+      "step": 2105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.992114100141241e-05,
+      "loss": 1.8562,
+      "step": 2106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.992106583549922e-05,
+      "loss": 2.0591,
+      "step": 2107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9920990633922138e-05,
+      "loss": 1.7829,
+      "step": 2108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.992091539668145e-05,
+      "loss": 1.8609,
+      "step": 2109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.992084012377741e-05,
+      "loss": 2.0125,
+      "step": 2110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9920764815210298e-05,
+      "loss": 2.1194,
+      "step": 2111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9920689470980382e-05,
+      "loss": 2.6981,
+      "step": 2112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9920614091087933e-05,
+      "loss": 1.9451,
+      "step": 2113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.992053867553322e-05,
+      "loss": 1.9954,
+      "step": 2114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9920463224316526e-05,
+      "loss": 2.0598,
+      "step": 2115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9920387737438106e-05,
+      "loss": 1.8481,
+      "step": 2116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9920312214898238e-05,
+      "loss": 2.0419,
+      "step": 2117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9920236656697195e-05,
+      "loss": 1.9923,
+      "step": 2118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9920161062835245e-05,
+      "loss": 2.0595,
+      "step": 2119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9920085433312665e-05,
+      "loss": 1.7545,
+      "step": 2120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.992000976812972e-05,
+      "loss": 1.952,
+      "step": 2121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.991993406728669e-05,
+      "loss": 1.8309,
+      "step": 2122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9919858330783842e-05,
+      "loss": 1.7641,
+      "step": 2123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.991978255862145e-05,
+      "loss": 1.7095,
+      "step": 2124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9919706750799784e-05,
+      "loss": 1.9489,
+      "step": 2125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.991963090731912e-05,
+      "loss": 1.7113,
+      "step": 2126
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9919555028179727e-05,
+      "loss": 1.895,
+      "step": 2127
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9919479113381883e-05,
+      "loss": 1.7713,
+      "step": 2128
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.991940316292585e-05,
+      "loss": 2.4979,
+      "step": 2129
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9919327176811915e-05,
+      "loss": 1.9653,
+      "step": 2130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9919251155040345e-05,
+      "loss": 1.9633,
+      "step": 2131
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9919175097611408e-05,
+      "loss": 1.8151,
+      "step": 2132
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9919099004525386e-05,
+      "loss": 1.5255,
+      "step": 2133
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.991902287578255e-05,
+      "loss": 1.778,
+      "step": 2134
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9918946711383168e-05,
+      "loss": 1.8882,
+      "step": 2135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.991887051132752e-05,
+      "loss": 1.9849,
+      "step": 2136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.991879427561588e-05,
+      "loss": 1.7459,
+      "step": 2137
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.991871800424852e-05,
+      "loss": 1.7751,
+      "step": 2138
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9918641697225716e-05,
+      "loss": 1.8147,
+      "step": 2139
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9918565354547738e-05,
+      "loss": 2.0301,
+      "step": 2140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9918488976214863e-05,
+      "loss": 2.0297,
+      "step": 2141
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.991841256222737e-05,
+      "loss": 1.8019,
+      "step": 2142
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9918336112585526e-05,
+      "loss": 1.985,
+      "step": 2143
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.991825962728961e-05,
+      "loss": 1.943,
+      "step": 2144
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9918183106339896e-05,
+      "loss": 1.8457,
+      "step": 2145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.991810654973666e-05,
+      "loss": 1.9366,
+      "step": 2146
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9918029957480173e-05,
+      "loss": 1.7637,
+      "step": 2147
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9917953329570717e-05,
+      "loss": 1.8693,
+      "step": 2148
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9917876666008565e-05,
+      "loss": 1.8655,
+      "step": 2149
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.991779996679399e-05,
+      "loss": 1.7945,
+      "step": 2150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.991772323192727e-05,
+      "loss": 1.8466,
+      "step": 2151
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.991764646140868e-05,
+      "loss": 2.084,
+      "step": 2152
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.99175696552385e-05,
+      "loss": 1.772,
+      "step": 2153
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9917492813417e-05,
+      "loss": 1.7669,
+      "step": 2154
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9917415935944455e-05,
+      "loss": 2.0276,
+      "step": 2155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9917339022821148e-05,
+      "loss": 1.6918,
+      "step": 2156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9917262074047355e-05,
+      "loss": 1.9878,
+      "step": 2157
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.991718508962335e-05,
+      "loss": 2.2508,
+      "step": 2158
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9917108069549406e-05,
+      "loss": 1.8336,
+      "step": 2159
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9917031013825808e-05,
+      "loss": 1.9695,
+      "step": 2160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9916953922452826e-05,
+      "loss": 1.846,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.991687679543074e-05,
+      "loss": 1.9016,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9916799632759832e-05,
+      "loss": 2.0606,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9916722434440368e-05,
+      "loss": 1.9391,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.991664520047264e-05,
+      "loss": 2.1304,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9916567930856908e-05,
+      "loss": 2.1045,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9916490625593467e-05,
+      "loss": 2.4067,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.991641328468258e-05,
+      "loss": 1.8135,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.991633590812454e-05,
+      "loss": 1.9188,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9916258495919613e-05,
+      "loss": 2.0317,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9916181048068085e-05,
+      "loss": 1.9826,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.991610356457023e-05,
+      "loss": 1.7939,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9916026045426324e-05,
+      "loss": 1.8912,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9915948490636653e-05,
+      "loss": 1.8192,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.991587090020149e-05,
+      "loss": 1.7888,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.991579327412112e-05,
+      "loss": 1.9452,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9915715612395816e-05,
+      "loss": 1.9085,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.991563791502586e-05,
+      "loss": 2.3553,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9915560182011534e-05,
+      "loss": 1.8708,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.991548241335311e-05,
+      "loss": 1.8407,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.991540460905087e-05,
+      "loss": 1.7722,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.99153267691051e-05,
+      "loss": 2.0805,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.991524889351607e-05,
+      "loss": 2.0032,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9915170982284068e-05,
+      "loss": 1.6548,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.991509303540937e-05,
+      "loss": 1.836,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.991501505289226e-05,
+      "loss": 1.8528,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9914937034733012e-05,
+      "loss": 1.7219,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9914858980931915e-05,
+      "loss": 1.5722,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.991478089148924e-05,
+      "loss": 2.1613,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9914702766405277e-05,
+      "loss": 1.8313,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.99146246056803e-05,
+      "loss": 1.859,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.991454640931459e-05,
+      "loss": 2.0782,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9914468177308434e-05,
+      "loss": 1.9543,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9914389909662106e-05,
+      "loss": 2.1231,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9914311606375895e-05,
+      "loss": 2.1196,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9914233267450074e-05,
+      "loss": 2.0639,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9914154892884935e-05,
+      "loss": 1.598,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9914076482680746e-05,
+      "loss": 2.1391,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9913998036837802e-05,
+      "loss": 1.9457,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.991391955535638e-05,
+      "loss": 1.9205,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9913841038236762e-05,
+      "loss": 1.7618,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9913762485479226e-05,
+      "loss": 1.7057,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9913683897084062e-05,
+      "loss": 2.0639,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9913605273051546e-05,
+      "loss": 1.6634,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9913526613381968e-05,
+      "loss": 1.9501,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9913447918075603e-05,
+      "loss": 1.7814,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9913369187132736e-05,
+      "loss": 2.2398,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9913290420553655e-05,
+      "loss": 1.9129,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9913211618338633e-05,
+      "loss": 2.0965,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9913132780487967e-05,
+      "loss": 1.9702,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.991305390700193e-05,
+      "loss": 2.2746,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9912974997880807e-05,
+      "loss": 2.2305,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.991289605312489e-05,
+      "loss": 1.6706,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9912817072734448e-05,
+      "loss": 1.8279,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9912738056709775e-05,
+      "loss": 1.9439,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9912659005051153e-05,
+      "loss": 1.7251,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.991257991775887e-05,
+      "loss": 2.08,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.99125007948332e-05,
+      "loss": 1.8165,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.875,
+      "learning_rate": 1.991242163627444e-05,
+      "loss": 2.0077,
+      "step": 2219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9912342442082867e-05,
+      "loss": 2.2214,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9912263212258763e-05,
+      "loss": 1.8373,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.991218394680242e-05,
+      "loss": 2.0992,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.991210464571412e-05,
+      "loss": 1.8675,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9912025308994146e-05,
+      "loss": 2.132,
+      "step": 2224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.991194593664279e-05,
+      "loss": 1.9322,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.991186652866033e-05,
+      "loss": 1.6247,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.991178708504705e-05,
+      "loss": 1.8999,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9911707605803243e-05,
+      "loss": 2.019,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.991162809092919e-05,
+      "loss": 1.699,
+      "step": 2229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.991154854042518e-05,
+      "loss": 1.9589,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9911468954291496e-05,
+      "loss": 2.2057,
+      "step": 2231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9911389332528425e-05,
+      "loss": 1.9789,
+      "step": 2232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9911309675136255e-05,
+      "loss": 1.8021,
+      "step": 2233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9911229982115267e-05,
+      "loss": 1.9579,
+      "step": 2234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9911150253465753e-05,
+      "loss": 1.7245,
+      "step": 2235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9911070489188e-05,
+      "loss": 1.9611,
+      "step": 2236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.991099068928229e-05,
+      "loss": 1.79,
+      "step": 2237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9910910853748912e-05,
+      "loss": 1.7574,
+      "step": 2238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9910830982588154e-05,
+      "loss": 1.666,
+      "step": 2239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9910751075800303e-05,
+      "loss": 1.9866,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9910671133385646e-05,
+      "loss": 1.7734,
+      "step": 2241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.991059115534447e-05,
+      "loss": 1.8264,
+      "step": 2242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.991051114167706e-05,
+      "loss": 2.0067,
+      "step": 2243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9910431092383708e-05,
+      "loss": 1.8525,
+      "step": 2244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9910351007464702e-05,
+      "loss": 2.0042,
+      "step": 2245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9910270886920324e-05,
+      "loss": 1.8893,
+      "step": 2246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.991019073075087e-05,
+      "loss": 1.9581,
+      "step": 2247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.991011053895662e-05,
+      "loss": 1.713,
+      "step": 2248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.991003031153787e-05,
+      "loss": 1.9075,
+      "step": 2249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.99099500484949e-05,
+      "loss": 2.2004,
+      "step": 2250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9909869749828008e-05,
+      "loss": 1.849,
+      "step": 2251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9909789415537478e-05,
+      "loss": 2.0206,
+      "step": 2252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0,
+      "learning_rate": 1.990970904562359e-05,
+      "loss": 2.0654,
+      "step": 2253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9909628640086654e-05,
+      "loss": 1.7205,
+      "step": 2254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.990954819892694e-05,
+      "loss": 2.2141,
+      "step": 2255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9909467722144745e-05,
+      "loss": 2.0548,
+      "step": 2256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9909387209740358e-05,
+      "loss": 1.9736,
+      "step": 2257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9909306661714064e-05,
+      "loss": 2.1394,
+      "step": 2258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9909226078066158e-05,
+      "loss": 1.8821,
+      "step": 2259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.990914545879693e-05,
+      "loss": 1.9257,
+      "step": 2260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9909064803906665e-05,
+      "loss": 1.7253,
+      "step": 2261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9908984113395653e-05,
+      "loss": 1.7264,
+      "step": 2262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.990890338726419e-05,
+      "loss": 1.992,
+      "step": 2263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9908822625512563e-05,
+      "loss": 1.7294,
+      "step": 2264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.990874182814106e-05,
+      "loss": 1.8897,
+      "step": 2265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9908660995149972e-05,
+      "loss": 1.9761,
+      "step": 2266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9908580126539597e-05,
+      "loss": 2.0481,
+      "step": 2267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9908499222310216e-05,
+      "loss": 1.533,
+      "step": 2268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9908418282462123e-05,
+      "loss": 2.0035,
+      "step": 2269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.990833730699561e-05,
+      "loss": 1.9503,
+      "step": 2270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9908256295910967e-05,
+      "loss": 1.927,
+      "step": 2271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.990817524920849e-05,
+      "loss": 2.1796,
+      "step": 2272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9908094166888462e-05,
+      "loss": 1.9538,
+      "step": 2273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.990801304895118e-05,
+      "loss": 2.0962,
+      "step": 2274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9907931895396934e-05,
+      "loss": 1.9199,
+      "step": 2275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9907850706226015e-05,
+      "loss": 1.7249,
+      "step": 2276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.990776948143872e-05,
+      "loss": 2.0512,
+      "step": 2277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 1.9907688221035334e-05,
+      "loss": 2.1059,
+      "step": 2278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9907606925016152e-05,
+      "loss": 1.8309,
+      "step": 2279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9907525593381464e-05,
+      "loss": 1.8613,
+      "step": 2280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9907444226131568e-05,
+      "loss": 1.7137,
+      "step": 2281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9907362823266752e-05,
+      "loss": 1.8967,
+      "step": 2282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9907281384787308e-05,
+      "loss": 2.1123,
+      "step": 2283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.990719991069353e-05,
+      "loss": 1.9868,
+      "step": 2284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9907118400985713e-05,
+      "loss": 1.7669,
+      "step": 2285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9907036855664148e-05,
+      "loss": 1.5766,
+      "step": 2286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.990695527472913e-05,
+      "loss": 1.8094,
+      "step": 2287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9906873658180948e-05,
+      "loss": 2.314,
+      "step": 2288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9906792006019902e-05,
+      "loss": 1.9366,
+      "step": 2289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9906710318246278e-05,
+      "loss": 1.9142,
+      "step": 2290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9906628594860377e-05,
+      "loss": 2.1029,
+      "step": 2291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9906546835862482e-05,
+      "loss": 1.8748,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.99064650412529e-05,
+      "loss": 1.7818,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.990638321103192e-05,
+      "loss": 2.0917,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9906301345199833e-05,
+      "loss": 2.0108,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9906219443756934e-05,
+      "loss": 2.0578,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.990613750670352e-05,
+      "loss": 1.9251,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9906055534039883e-05,
+      "loss": 1.7833,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.990597352576632e-05,
+      "loss": 2.2029,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9905891481883126e-05,
+      "loss": 2.1103,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.990580940239059e-05,
+      "loss": 1.6898,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.990572728728902e-05,
+      "loss": 1.8899,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9905645136578697e-05,
+      "loss": 1.6375,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9905562950259925e-05,
+      "loss": 1.9967,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9905480728333e-05,
+      "loss": 1.4698,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9905398470798207e-05,
+      "loss": 2.0085,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9905316177655855e-05,
+      "loss": 2.2094,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.990523384890623e-05,
+      "loss": 1.7522,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9905151484549636e-05,
+      "loss": 1.9177,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.990506908458636e-05,
+      "loss": 1.8299,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9904986649016704e-05,
+      "loss": 2.1438,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9904904177840965e-05,
+      "loss": 1.8944,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9904821671059437e-05,
+      "loss": 2.0736,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.990473912867242e-05,
+      "loss": 1.9588,
+      "step": 2314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9904656550680204e-05,
+      "loss": 1.8633,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9904573937083095e-05,
+      "loss": 1.8324,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.990449128788138e-05,
+      "loss": 1.7936,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9904408603075363e-05,
+      "loss": 1.9258,
+      "step": 2318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.990432588266534e-05,
+      "loss": 1.9609,
+      "step": 2319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.990424312665161e-05,
+      "loss": 1.8828,
+      "step": 2320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9904160335034466e-05,
+      "loss": 2.0187,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9904077507814207e-05,
+      "loss": 1.7712,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9903994644991135e-05,
+      "loss": 2.025,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.990391174656554e-05,
+      "loss": 1.771,
+      "step": 2324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9903828812537728e-05,
+      "loss": 1.8398,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9903745842907992e-05,
+      "loss": 1.7075,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.990366283767663e-05,
+      "loss": 1.534,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9903579796843948e-05,
+      "loss": 2.1131,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.990349672041023e-05,
+      "loss": 2.3812,
+      "step": 2329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.990341360837579e-05,
+      "loss": 1.9184,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.990333046074092e-05,
+      "loss": 1.8694,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9903247277505918e-05,
+      "loss": 1.956,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9903164058671083e-05,
+      "loss": 2.0095,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9903080804236717e-05,
+      "loss": 1.9918,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9902997514203115e-05,
+      "loss": 1.9322,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.990291418857058e-05,
+      "loss": 1.9157,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.990283082733941e-05,
+      "loss": 1.7924,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9902747430509907e-05,
+      "loss": 1.7361,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25,
+      "learning_rate": 1.990266399808237e-05,
+      "loss": 1.5449,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9902580530057094e-05,
+      "loss": 1.9556,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9902497026434387e-05,
+      "loss": 2.0115,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.990241348721454e-05,
+      "loss": 1.9949,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9902329912397863e-05,
+      "loss": 2.0729,
+      "step": 2343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.990224630198465e-05,
+      "loss": 2.0167,
+      "step": 2344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9902162655975203e-05,
+      "loss": 1.7695,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9902078974369823e-05,
+      "loss": 1.3012,
+      "step": 2346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9901995257168808e-05,
+      "loss": 1.9296,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9901911504372467e-05,
+      "loss": 1.842,
+      "step": 2348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9901827715981094e-05,
+      "loss": 1.686,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.990174389199499e-05,
+      "loss": 1.9333,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.9901660032414464e-05,
+      "loss": 1.2857,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9901576137239806e-05,
+      "loss": 1.5881,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9901492206471325e-05,
+      "loss": 2.0639,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.9901408240109325e-05,
+      "loss": 2.386,
+      "step": 2354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9901324238154102e-05,
+      "loss": 2.268,
+      "step": 2355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9901240200605955e-05,
+      "loss": 1.7371,
+      "step": 2356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9901156127465195e-05,
+      "loss": 2.0943,
+      "step": 2357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9901072018732122e-05,
+      "loss": 1.9922,
+      "step": 2358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9900987874407035e-05,
+      "loss": 1.9728,
+      "step": 2359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9900903694490238e-05,
+      "loss": 1.9018,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.990081947898203e-05,
+      "loss": 2.0309,
+      "step": 2361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9900735227882722e-05,
+      "loss": 2.0642,
+      "step": 2362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9900650941192613e-05,
+      "loss": 1.9081,
+      "step": 2363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9900566618912003e-05,
+      "loss": 2.0327,
+      "step": 2364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.99004822610412e-05,
+      "loss": 1.7158,
+      "step": 2365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.99003978675805e-05,
+      "loss": 2.2607,
+      "step": 2366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9900313438530213e-05,
+      "loss": 1.9471,
+      "step": 2367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9900228973890642e-05,
+      "loss": 2.0557,
+      "step": 2368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9900144473662088e-05,
+      "loss": 1.8883,
+      "step": 2369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9900059937844858e-05,
+      "loss": 2.116,
+      "step": 2370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9899975366439254e-05,
+      "loss": 1.9114,
+      "step": 2371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9899890759445575e-05,
+      "loss": 2.0208,
+      "step": 2372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9899806116864135e-05,
+      "loss": 1.8786,
+      "step": 2373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.989972143869523e-05,
+      "loss": 1.746,
+      "step": 2374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.989963672493917e-05,
+      "loss": 1.9048,
+      "step": 2375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9899551975596253e-05,
+      "loss": 1.7675,
+      "step": 2376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9899467190666792e-05,
+      "loss": 1.6159,
+      "step": 2377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.989938237015109e-05,
+      "loss": 2.0852,
+      "step": 2378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.989929751404944e-05,
+      "loss": 2.1833,
+      "step": 2379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9899212622362168e-05,
+      "loss": 1.9123,
+      "step": 2380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9899127695089563e-05,
+      "loss": 1.9676,
+      "step": 2381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.989904273223193e-05,
+      "loss": 2.0482,
+      "step": 2382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9898957733789585e-05,
+      "loss": 2.049,
+      "step": 2383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.989887269976283e-05,
+      "loss": 1.9929,
+      "step": 2384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.989878763015196e-05,
+      "loss": 2.2408,
+      "step": 2385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9898702524957297e-05,
+      "loss": 1.9906,
+      "step": 2386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9898617384179138e-05,
+      "loss": 2.4668,
+      "step": 2387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.989853220781779e-05,
+      "loss": 1.8708,
+      "step": 2388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9898446995873558e-05,
+      "loss": 2.0574,
+      "step": 2389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9898361748346754e-05,
+      "loss": 1.9407,
+      "step": 2390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9898276465237674e-05,
+      "loss": 1.8429,
+      "step": 2391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9898191146546638e-05,
+      "loss": 2.1136,
+      "step": 2392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.989810579227394e-05,
+      "loss": 1.8868,
+      "step": 2393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9898020402419895e-05,
+      "loss": 1.9121,
+      "step": 2394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.989793497698481e-05,
+      "loss": 1.8945,
+      "step": 2395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9897849515968987e-05,
+      "loss": 1.8204,
+      "step": 2396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9897764019372734e-05,
+      "loss": 1.9717,
+      "step": 2397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.989767848719636e-05,
+      "loss": 1.8087,
+      "step": 2398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9897592919440175e-05,
+      "loss": 1.5329,
+      "step": 2399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9897507316104486e-05,
+      "loss": 2.1968,
+      "step": 2400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9897421677189595e-05,
+      "loss": 2.1761,
+      "step": 2401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9897336002695816e-05,
+      "loss": 1.6996,
+      "step": 2402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9897250292623452e-05,
+      "loss": 2.0475,
+      "step": 2403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9897164546972818e-05,
+      "loss": 1.9888,
+      "step": 2404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9897078765744215e-05,
+      "loss": 1.9463,
+      "step": 2405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9896992948937957e-05,
+      "loss": 1.9715,
+      "step": 2406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9896907096554348e-05,
+      "loss": 2.0192,
+      "step": 2407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9896821208593698e-05,
+      "loss": 1.953,
+      "step": 2408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9896735285056315e-05,
+      "loss": 2.0878,
+      "step": 2409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9896649325942513e-05,
+      "loss": 1.8541,
+      "step": 2410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9896563331252597e-05,
+      "loss": 1.9457,
+      "step": 2411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.989647730098687e-05,
+      "loss": 2.1043,
+      "step": 2412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9896391235145657e-05,
+      "loss": 2.0905,
+      "step": 2413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.989630513372925e-05,
+      "loss": 1.8733,
+      "step": 2414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.989621899673797e-05,
+      "loss": 2.0623,
+      "step": 2415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9896132824172123e-05,
+      "loss": 1.7117,
+      "step": 2416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.989604661603202e-05,
+      "loss": 1.8924,
+      "step": 2417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9895960372317965e-05,
+      "loss": 2.1624,
+      "step": 2418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.989587409303028e-05,
+      "loss": 1.7714,
+      "step": 2419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.989578777816926e-05,
+      "loss": 1.9381,
+      "step": 2420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.989570142773523e-05,
+      "loss": 1.7574,
+      "step": 2421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.989561504172849e-05,
+      "loss": 1.8107,
+      "step": 2422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9895528620149354e-05,
+      "loss": 1.8295,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9895442162998136e-05,
+      "loss": 2.3299,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.989535567027514e-05,
+      "loss": 1.7942,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9895269141980683e-05,
+      "loss": 2.1548,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9895182578115073e-05,
+      "loss": 1.6862,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.989509597867862e-05,
+      "loss": 1.9933,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9895009343671638e-05,
+      "loss": 1.8797,
+      "step": 2429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.989492267309444e-05,
+      "loss": 1.7734,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.989483596694733e-05,
+      "loss": 2.0965,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9894749225230633e-05,
+      "loss": 1.8007,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9894662447944647e-05,
+      "loss": 2.3603,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9894575635089687e-05,
+      "loss": 1.8252,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5,
+      "learning_rate": 1.989448878666607e-05,
+      "loss": 2.0872,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.989440190267411e-05,
+      "loss": 1.9187,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.989431498311411e-05,
+      "loss": 2.234,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9894228027986392e-05,
+      "loss": 1.854,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9894141037291258e-05,
+      "loss": 2.1263,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.989405401102903e-05,
+      "loss": 1.9717,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9893966949200018e-05,
+      "loss": 1.8449,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9893879851804537e-05,
+      "loss": 1.6945,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9893792718842894e-05,
+      "loss": 1.8612,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9893705550315404e-05,
+      "loss": 1.856,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9893618346222387e-05,
+      "loss": 1.9167,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.989353110656415e-05,
+      "loss": 1.722,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.989344383134101e-05,
+      "loss": 2.0644,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9893356520553277e-05,
+      "loss": 1.6956,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9893269174201267e-05,
+      "loss": 2.2616,
+      "step": 2449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9893181792285297e-05,
+      "loss": 2.0566,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9893094374805673e-05,
+      "loss": 1.8185,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.989300692176272e-05,
+      "loss": 1.6324,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9892919433156745e-05,
+      "loss": 1.7161,
+      "step": 2453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9892831908988063e-05,
+      "loss": 2.0231,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.989274434925699e-05,
+      "loss": 1.6407,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9892656753963843e-05,
+      "loss": 1.9584,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9892569123108933e-05,
+      "loss": 1.7058,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9892481456692577e-05,
+      "loss": 1.6955,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9892393754715088e-05,
+      "loss": 1.8472,
+      "step": 2459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.989230601717678e-05,
+      "loss": 1.6939,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.989221824407798e-05,
+      "loss": 1.8878,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9892130435418985e-05,
+      "loss": 1.4452,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9892042591200127e-05,
+      "loss": 1.8071,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.989195471142171e-05,
+      "loss": 1.7322,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9891866796084063e-05,
+      "loss": 1.9829,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.989177884518749e-05,
+      "loss": 2.0536,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9891690858732313e-05,
+      "loss": 1.6166,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9891602836718844e-05,
+      "loss": 2.1565,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9891514779147403e-05,
+      "loss": 1.8872,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9891426686018308e-05,
+      "loss": 1.7026,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9891338557331868e-05,
+      "loss": 2.0878,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9891250393088408e-05,
+      "loss": 1.7778,
+      "step": 2472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9891162193288243e-05,
+      "loss": 1.6712,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9891073957931687e-05,
+      "loss": 1.7967,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9890985687019064e-05,
+      "loss": 2.0291,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.989089738055068e-05,
+      "loss": 1.9546,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9890809038526867e-05,
+      "loss": 1.691,
+      "step": 2477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9890720660947927e-05,
+      "loss": 1.7757,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9890632247814186e-05,
+      "loss": 1.8435,
+      "step": 2479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9890543799125966e-05,
+      "loss": 1.9011,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.989045531488358e-05,
+      "loss": 1.7314,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9890366795087343e-05,
+      "loss": 2.0048,
+      "step": 2482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9890278239737576e-05,
+      "loss": 1.7051,
+      "step": 2483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9890189648834598e-05,
+      "loss": 2.9052,
+      "step": 2484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.989010102237873e-05,
+      "loss": 1.9205,
+      "step": 2485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9890012360370286e-05,
+      "loss": 1.8139,
+      "step": 2486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9889923662809584e-05,
+      "loss": 1.8718,
+      "step": 2487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9889834929696946e-05,
+      "loss": 1.6193,
+      "step": 2488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9889746161032696e-05,
+      "loss": 2.0157,
+      "step": 2489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.988965735681714e-05,
+      "loss": 2.1889,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.988956851705061e-05,
+      "loss": 2.0228,
+      "step": 2491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9889479641733417e-05,
+      "loss": 1.9188,
+      "step": 2492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9889390730865884e-05,
+      "loss": 1.8542,
+      "step": 2493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.988930178444833e-05,
+      "loss": 1.5444,
+      "step": 2494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9889212802481076e-05,
+      "loss": 1.6486,
+      "step": 2495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.988912378496444e-05,
+      "loss": 2.1706,
+      "step": 2496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9889034731898746e-05,
+      "loss": 2.0776,
+      "step": 2497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9888945643284307e-05,
+      "loss": 1.9634,
+      "step": 2498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9888856519121447e-05,
+      "loss": 1.8621,
+      "step": 2499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9888767359410487e-05,
+      "loss": 2.4233,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9888678164151747e-05,
+      "loss": 1.9144,
+      "step": 2501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.988858893334555e-05,
+      "loss": 1.6615,
+      "step": 2502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.988849966699221e-05,
+      "loss": 2.2684,
+      "step": 2503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9888410365092056e-05,
+      "loss": 1.7674,
+      "step": 2504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9888321027645402e-05,
+      "loss": 1.9915,
+      "step": 2505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9888231654652573e-05,
+      "loss": 1.4668,
+      "step": 2506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.988814224611389e-05,
+      "loss": 2.0648,
+      "step": 2507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.988805280202968e-05,
+      "loss": 1.9328,
+      "step": 2508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.988796332240025e-05,
+      "loss": 1.7582,
+      "step": 2509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9887873807225935e-05,
+      "loss": 1.9064,
+      "step": 2510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.988778425650705e-05,
+      "loss": 1.6948,
+      "step": 2511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.988769467024392e-05,
+      "loss": 1.7125,
+      "step": 2512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9887605048436867e-05,
+      "loss": 1.8079,
+      "step": 2513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.988751539108621e-05,
+      "loss": 1.8935,
+      "step": 2514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9887425698192274e-05,
+      "loss": 1.953,
+      "step": 2515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9887335969755384e-05,
+      "loss": 2.3496,
+      "step": 2516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9887246205775854e-05,
+      "loss": 1.9702,
+      "step": 2517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9887156406254018e-05,
+      "loss": 1.8548,
+      "step": 2518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.988706657119019e-05,
+      "loss": 1.7821,
+      "step": 2519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9886976700584693e-05,
+      "loss": 1.9869,
+      "step": 2520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9886886794437855e-05,
+      "loss": 2.0919,
+      "step": 2521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9886796852750003e-05,
+      "loss": 1.9481,
+      "step": 2522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9886706875521447e-05,
+      "loss": 1.845,
+      "step": 2523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.988661686275252e-05,
+      "loss": 2.1893,
+      "step": 2524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9886526814443546e-05,
+      "loss": 2.0686,
+      "step": 2525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9886436730594843e-05,
+      "loss": 1.6037,
+      "step": 2526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.988634661120674e-05,
+      "loss": 1.8775,
+      "step": 2527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9886256456279558e-05,
+      "loss": 1.8469,
+      "step": 2528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9886166265813623e-05,
+      "loss": 1.7274,
+      "step": 2529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9886076039809258e-05,
+      "loss": 1.8134,
+      "step": 2530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.988598577826679e-05,
+      "loss": 1.793,
+      "step": 2531
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9885895481186534e-05,
+      "loss": 2.6422,
+      "step": 2532
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.988580514856883e-05,
+      "loss": 1.6254,
+      "step": 2533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9885714780413986e-05,
+      "loss": 1.8933,
+      "step": 2534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9885624376722342e-05,
+      "loss": 1.9381,
+      "step": 2535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.988553393749421e-05,
+      "loss": 1.9092,
+      "step": 2536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9885443462729927e-05,
+      "loss": 1.8649,
+      "step": 2537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9885352952429812e-05,
+      "loss": 2.063,
+      "step": 2538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.988526240659419e-05,
+      "loss": 1.993,
+      "step": 2539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.988517182522339e-05,
+      "loss": 1.8632,
+      "step": 2540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.988508120831773e-05,
+      "loss": 1.802,
+      "step": 2541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9884990555877545e-05,
+      "loss": 1.6367,
+      "step": 2542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9884899867903152e-05,
+      "loss": 1.4267,
+      "step": 2543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9884809144394887e-05,
+      "loss": 2.0087,
+      "step": 2544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9884718385353068e-05,
+      "loss": 1.8188,
+      "step": 2545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9884627590778023e-05,
+      "loss": 1.7094,
+      "step": 2546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9884536760670082e-05,
+      "loss": 1.9891,
+      "step": 2547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9884445895029566e-05,
+      "loss": 1.7898,
+      "step": 2548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9884354993856808e-05,
+      "loss": 1.6909,
+      "step": 2549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9884264057152132e-05,
+      "loss": 1.6506,
+      "step": 2550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.988417308491586e-05,
+      "loss": 1.6801,
+      "step": 2551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9884082077148326e-05,
+      "loss": 2.043,
+      "step": 2552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9883991033849856e-05,
+      "loss": 2.0538,
+      "step": 2553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.988389995502077e-05,
+      "loss": 2.1898,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9883808840661406e-05,
+      "loss": 1.7814,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9883717690772086e-05,
+      "loss": 1.5071,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9883626505353136e-05,
+      "loss": 2.0873,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9883535284404888e-05,
+      "loss": 1.7928,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.988344402792767e-05,
+      "loss": 1.8466,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9883352735921807e-05,
+      "loss": 1.7047,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9883261408387625e-05,
+      "loss": 1.7296,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.988317004532546e-05,
+      "loss": 1.836,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9883078646735635e-05,
+      "loss": 1.9385,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9882987212618477e-05,
+      "loss": 1.973,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9882895742974318e-05,
+      "loss": 1.6264,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9882804237803487e-05,
+      "loss": 1.6642,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.988271269710631e-05,
+      "loss": 1.7216,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.988262112088312e-05,
+      "loss": 1.9623,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9882529509134246e-05,
+      "loss": 2.0729,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9882437861860014e-05,
+      "loss": 1.8629,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9882346179060753e-05,
+      "loss": 2.061,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9882254460736797e-05,
+      "loss": 2.1566,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9882162706888473e-05,
+      "loss": 1.9057,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.988207091751611e-05,
+      "loss": 2.7229,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.988197909262004e-05,
+      "loss": 1.9873,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9881887232200592e-05,
+      "loss": 2.1143,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9881795336258097e-05,
+      "loss": 1.8889,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9881703404792883e-05,
+      "loss": 1.9269,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9881611437805285e-05,
+      "loss": 2.1643,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9881519435295627e-05,
+      "loss": 1.841,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9881427397264246e-05,
+      "loss": 1.8861,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9881335323711467e-05,
+      "loss": 2.0696,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.988124321463763e-05,
+      "loss": 2.0263,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9881151070043056e-05,
+      "loss": 2.0324,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9881058889928085e-05,
+      "loss": 2.0459,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.988096667429304e-05,
+      "loss": 1.776,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.988087442313826e-05,
+      "loss": 2.0876,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.988078213646407e-05,
+      "loss": 1.8584,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9880689814270808e-05,
+      "loss": 2.5693,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.98805974565588e-05,
+      "loss": 1.8296,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9880505063328385e-05,
+      "loss": 1.9594,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9880412634579886e-05,
+      "loss": 1.8107,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.988032017031364e-05,
+      "loss": 1.982,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9880227670529987e-05,
+      "loss": 1.9778,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9880135135229242e-05,
+      "loss": 1.8237,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9880042564411755e-05,
+      "loss": 1.7144,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.987994995807785e-05,
+      "loss": 2.1122,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.9879857316227863e-05,
+      "loss": 2.2831,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.987976463886212e-05,
+      "loss": 1.8771,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9879671925980964e-05,
+      "loss": 2.0679,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.987957917758472e-05,
+      "loss": 1.7419,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.987948639367373e-05,
+      "loss": 1.7565,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.987939357424832e-05,
+      "loss": 1.8527,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.987930071930883e-05,
+      "loss": 2.0894,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9879207828855586e-05,
+      "loss": 1.5893,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9879114902888928e-05,
+      "loss": 2.0342,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.987902194140919e-05,
+      "loss": 1.6635,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.98789289444167e-05,
+      "loss": 1.9498,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.98788359119118e-05,
+      "loss": 1.7828,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.987874284389482e-05,
+      "loss": 1.9806,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.98786497403661e-05,
+      "loss": 1.8864,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9878556601325965e-05,
+      "loss": 1.745,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9878463426774757e-05,
+      "loss": 1.7135,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9878370216712808e-05,
+      "loss": 1.8194,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9878276971140455e-05,
+      "loss": 1.8091,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9878183690058036e-05,
+      "loss": 1.8971,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9878090373465877e-05,
+      "loss": 1.9074,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9877997021364323e-05,
+      "loss": 2.0463,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9877903633753703e-05,
+      "loss": 2.0052,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.987781021063436e-05,
+      "loss": 2.2054,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9877716752006623e-05,
+      "loss": 1.5904,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.987762325787083e-05,
+      "loss": 1.7685,
+      "step": 2622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9877529728227313e-05,
+      "loss": 2.056,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.987743616307642e-05,
+      "loss": 2.1025,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.9877342562418474e-05,
+      "loss": 1.9629,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9877248926253823e-05,
+      "loss": 1.9743,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9877155254582793e-05,
+      "loss": 1.6957,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9877061547405728e-05,
+      "loss": 2.106,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9876967804722963e-05,
+      "loss": 1.4959,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9876874026534833e-05,
+      "loss": 1.9668,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9876780212841678e-05,
+      "loss": 2.1323,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9876686363643832e-05,
+      "loss": 1.9069,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9876592478941638e-05,
+      "loss": 1.7911,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9876498558735427e-05,
+      "loss": 1.6425,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.987640460302554e-05,
+      "loss": 1.7233,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9876310611812317e-05,
+      "loss": 1.9553,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9876216585096087e-05,
+      "loss": 1.7408,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.98761225228772e-05,
+      "loss": 1.3873,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9876028425155985e-05,
+      "loss": 1.8414,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9875934291932782e-05,
+      "loss": 2.0161,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9875840123207935e-05,
+      "loss": 1.9501,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9875745918981776e-05,
+      "loss": 1.606,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9875651679254643e-05,
+      "loss": 2.1792,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.987555740402688e-05,
+      "loss": 1.8705,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9875463093298828e-05,
+      "loss": 2.1034,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9875368747070817e-05,
+      "loss": 2.0777,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.987527436534319e-05,
+      "loss": 1.7571,
+      "step": 2647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9875179948116292e-05,
+      "loss": 1.7472,
+      "step": 2648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9875085495390453e-05,
+      "loss": 1.9132,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.987499100716602e-05,
+      "loss": 2.1265,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9874896483443327e-05,
+      "loss": 2.2558,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.987480192422272e-05,
+      "loss": 1.8729,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.987470732950453e-05,
+      "loss": 1.8024,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9874612699289107e-05,
+      "loss": 2.1282,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9874518033576783e-05,
+      "loss": 1.7936,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9874423332367906e-05,
+      "loss": 1.89,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9874328595662812e-05,
+      "loss": 1.8663,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.987423382346184e-05,
+      "loss": 2.1045,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9874139015765333e-05,
+      "loss": 2.0886,
+      "step": 2659
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9874044172573632e-05,
+      "loss": 2.0963,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9873949293887077e-05,
+      "loss": 2.0689,
+      "step": 2661
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.987385437970601e-05,
+      "loss": 2.0274,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.9873759430030767e-05,
+      "loss": 2.1917,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.98736644448617e-05,
+      "loss": 1.7907,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.987356942419914e-05,
+      "loss": 2.1813,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9873474368043436e-05,
+      "loss": 1.8833,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9873379276394926e-05,
+      "loss": 2.0093,
+      "step": 2667
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9873284149253952e-05,
+      "loss": 2.0529,
+      "step": 2668
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9873188986620855e-05,
+      "loss": 1.9859,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9873093788495983e-05,
+      "loss": 1.7038,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.987299855487967e-05,
+      "loss": 1.8482,
+      "step": 2671
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9872903285772263e-05,
+      "loss": 2.0995,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9872807981174102e-05,
+      "loss": 1.6297,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9872712641085532e-05,
+      "loss": 2.0172,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.98726172655069e-05,
+      "loss": 1.8683,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9872521854438537e-05,
+      "loss": 1.7912,
+      "step": 2676
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9872426407880795e-05,
+      "loss": 1.9807,
+      "step": 2677
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9872330925834013e-05,
+      "loss": 2.1513,
+      "step": 2678
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.987223540829854e-05,
+      "loss": 1.7385,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9872139855274712e-05,
+      "loss": 1.8559,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.987204426676288e-05,
+      "loss": 1.8661,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.987194864276338e-05,
+      "loss": 1.9748,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9871852983276558e-05,
+      "loss": 1.9458,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9871757288302762e-05,
+      "loss": 1.8869,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.987166155784233e-05,
+      "loss": 1.6112,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9871565791895614e-05,
+      "loss": 2.0209,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9871469990462953e-05,
+      "loss": 2.161,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9871374153544688e-05,
+      "loss": 2.16,
+      "step": 2688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9871278281141173e-05,
+      "loss": 2.0431,
+      "step": 2689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.987118237325274e-05,
+      "loss": 1.8902,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9871086429879746e-05,
+      "loss": 1.6799,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.987099045102253e-05,
+      "loss": 1.7722,
+      "step": 2692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.9870894436681432e-05,
+      "loss": 1.7863,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.987079838685681e-05,
+      "loss": 1.7773,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9870702301549e-05,
+      "loss": 1.977,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.987060618075835e-05,
+      "loss": 1.9158,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9870510024485205e-05,
+      "loss": 1.5893,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9870413832729913e-05,
+      "loss": 2.1256,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9870317605492813e-05,
+      "loss": 2.0102,
+      "step": 2699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.987022134277426e-05,
+      "loss": 2.0422,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.987012504457459e-05,
+      "loss": 1.8054,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9870028710894157e-05,
+      "loss": 2.0063,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9869932341733307e-05,
+      "loss": 1.7125,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9869835937092383e-05,
+      "loss": 1.9169,
+      "step": 2704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9869739496971734e-05,
+      "loss": 1.8861,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9869643021371706e-05,
+      "loss": 1.901,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9869546510292644e-05,
+      "loss": 1.907,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9869449963734894e-05,
+      "loss": 1.8381,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9869353381698807e-05,
+      "loss": 1.7971,
+      "step": 2709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9869256764184732e-05,
+      "loss": 1.8983,
+      "step": 2710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.986916011119301e-05,
+      "loss": 2.2578,
+      "step": 2711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.986906342272399e-05,
+      "loss": 2.1433,
+      "step": 2712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9868966698778023e-05,
+      "loss": 1.7193,
+      "step": 2713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.9868869939355456e-05,
+      "loss": 3.5369,
+      "step": 2714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9868773144456632e-05,
+      "loss": 1.9472,
+      "step": 2715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9868676314081907e-05,
+      "loss": 1.8967,
+      "step": 2716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9868579448231622e-05,
+      "loss": 1.8343,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9868482546906127e-05,
+      "loss": 2.0458,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.986838561010577e-05,
+      "loss": 1.8153,
+      "step": 2719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9868288637830905e-05,
+      "loss": 2.0192,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9868191630081873e-05,
+      "loss": 2.0352,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9868094586859026e-05,
+      "loss": 1.8228,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9867997508162714e-05,
+      "loss": 2.0055,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9867900393993283e-05,
+      "loss": 1.8134,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9867803244351083e-05,
+      "loss": 2.0842,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9867706059236466e-05,
+      "loss": 1.8638,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.986760883864978e-05,
+      "loss": 1.5413,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.986751158259137e-05,
+      "loss": 2.0875,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9867414291061595e-05,
+      "loss": 1.9329,
+      "step": 2729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9867316964060794e-05,
+      "loss": 1.6616,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.986721960158933e-05,
+      "loss": 1.7157,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9867122203647536e-05,
+      "loss": 2.204,
+      "step": 2732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9867024770235774e-05,
+      "loss": 2.1113,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9866927301354392e-05,
+      "loss": 1.7336,
+      "step": 2734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.986682979700374e-05,
+      "loss": 2.0585,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9866732257184168e-05,
+      "loss": 1.6508,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9866634681896027e-05,
+      "loss": 1.9859,
+      "step": 2737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.986653707113967e-05,
+      "loss": 1.7322,
+      "step": 2738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.986643942491544e-05,
+      "loss": 2.1256,
+      "step": 2739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.98663417432237e-05,
+      "loss": 2.0579,
+      "step": 2740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.986624402606479e-05,
+      "loss": 1.8805,
+      "step": 2741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.986614627343907e-05,
+      "loss": 1.7974,
+      "step": 2742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9866048485346884e-05,
+      "loss": 1.8867,
+      "step": 2743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.986595066178859e-05,
+      "loss": 1.7172,
+      "step": 2744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9865852802764533e-05,
+      "loss": 1.6637,
+      "step": 2745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.986575490827507e-05,
+      "loss": 1.8632,
+      "step": 2746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9865656978320552e-05,
+      "loss": 2.2104,
+      "step": 2747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.986555901290133e-05,
+      "loss": 1.7548,
+      "step": 2748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9865461012017756e-05,
+      "loss": 2.1157,
+      "step": 2749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9865362975670185e-05,
+      "loss": 1.9027,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9865264903858964e-05,
+      "loss": 1.7347,
+      "step": 2751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.986516679658445e-05,
+      "loss": 1.727,
+      "step": 2752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9865068653846995e-05,
+      "loss": 1.8396,
+      "step": 2753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.986497047564695e-05,
+      "loss": 2.3209,
+      "step": 2754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9864872261984672e-05,
+      "loss": 2.0413,
+      "step": 2755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9864774012860507e-05,
+      "loss": 1.8783,
+      "step": 2756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9864675728274813e-05,
+      "loss": 2.0662,
+      "step": 2757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9864577408227946e-05,
+      "loss": 1.864,
+      "step": 2758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9864479052720254e-05,
+      "loss": 1.9734,
+      "step": 2759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.986438066175209e-05,
+      "loss": 1.8039,
+      "step": 2760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9864282235323818e-05,
+      "loss": 2.1502,
+      "step": 2761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9864183773435778e-05,
+      "loss": 1.9068,
+      "step": 2762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9864085276088333e-05,
+      "loss": 1.5402,
+      "step": 2763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.986398674328183e-05,
+      "loss": 2.1734,
+      "step": 2764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9863888175016633e-05,
+      "loss": 2.2467,
+      "step": 2765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9863789571293085e-05,
+      "loss": 1.8923,
+      "step": 2766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.986369093211155e-05,
+      "loss": 1.7923,
+      "step": 2767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9863592257472377e-05,
+      "loss": 1.6225,
+      "step": 2768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9863493547375924e-05,
+      "loss": 2.2534,
+      "step": 2769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9863394801822544e-05,
+      "loss": 1.885,
+      "step": 2770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.986329602081259e-05,
+      "loss": 1.9109,
+      "step": 2771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9863197204346426e-05,
+      "loss": 1.8922,
+      "step": 2772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9863098352424396e-05,
+      "loss": 2.1484,
+      "step": 2773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9862999465046858e-05,
+      "loss": 1.9745,
+      "step": 2774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.986290054221417e-05,
+      "loss": 1.95,
+      "step": 2775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.986280158392669e-05,
+      "loss": 1.9408,
+      "step": 2776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9862702590184772e-05,
+      "loss": 1.7976,
+      "step": 2777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9862603560988766e-05,
+      "loss": 2.13,
+      "step": 2778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9862504496339036e-05,
+      "loss": 1.9149,
+      "step": 2779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9862405396235933e-05,
+      "loss": 2.0735,
+      "step": 2780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9862306260679815e-05,
+      "loss": 1.8464,
+      "step": 2781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.986220708967104e-05,
+      "loss": 1.8819,
+      "step": 2782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.986210788320996e-05,
+      "loss": 2.2926,
+      "step": 2783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9862008641296938e-05,
+      "loss": 1.87,
+      "step": 2784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9861909363932325e-05,
+      "loss": 2.0714,
+      "step": 2785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.986181005111648e-05,
+      "loss": 1.6054,
+      "step": 2786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9861710702849763e-05,
+      "loss": 2.2613,
+      "step": 2787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9861611319132524e-05,
+      "loss": 1.8084,
+      "step": 2788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.986151189996513e-05,
+      "loss": 2.0122,
+      "step": 2789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9861412445347927e-05,
+      "loss": 2.2479,
+      "step": 2790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9861312955281283e-05,
+      "loss": 2.0298,
+      "step": 2791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.986121342976555e-05,
+      "loss": 1.8629,
+      "step": 2792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.986111386880108e-05,
+      "loss": 1.8145,
+      "step": 2793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9861014272388245e-05,
+      "loss": 1.9671,
+      "step": 2794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9860914640527395e-05,
+      "loss": 2.0256,
+      "step": 2795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9860814973218887e-05,
+      "loss": 1.8857,
+      "step": 2796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9860715270463084e-05,
+      "loss": 2.0074,
+      "step": 2797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9860615532260337e-05,
+      "loss": 1.7458,
+      "step": 2798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.9860515758611014e-05,
+      "loss": 2.8696,
+      "step": 2799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9860415949515465e-05,
+      "loss": 2.0674,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9860316104974058e-05,
+      "loss": 1.64,
+      "step": 2801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.986021622498714e-05,
+      "loss": 2.0377,
+      "step": 2802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.375,
+      "learning_rate": 1.986011630955508e-05,
+      "loss": 2.0095,
+      "step": 2803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9860016358678233e-05,
+      "loss": 2.0082,
+      "step": 2804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.985991637235696e-05,
+      "loss": 1.8485,
+      "step": 2805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.985981635059162e-05,
+      "loss": 1.8119,
+      "step": 2806
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.985971629338257e-05,
+      "loss": 2.1362,
+      "step": 2807
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9859616200730172e-05,
+      "loss": 1.7409,
+      "step": 2808
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9859516072634785e-05,
+      "loss": 2.3306,
+      "step": 2809
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9859415909096772e-05,
+      "loss": 1.5952,
+      "step": 2810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9859315710116488e-05,
+      "loss": 2.0459,
+      "step": 2811
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9859215475694295e-05,
+      "loss": 1.8546,
+      "step": 2812
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9859115205830557e-05,
+      "loss": 2.2225,
+      "step": 2813
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9859014900525628e-05,
+      "loss": 1.9254,
+      "step": 2814
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9858914559779876e-05,
+      "loss": 2.0784,
+      "step": 2815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9858814183593655e-05,
+      "loss": 1.7356,
+      "step": 2816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9858713771967328e-05,
+      "loss": 1.7618,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9858613324901258e-05,
+      "loss": 1.8831,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9858512842395804e-05,
+      "loss": 2.2248,
+      "step": 2819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9858412324451328e-05,
+      "loss": 2.1102,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9858311771068193e-05,
+      "loss": 1.4833,
+      "step": 2821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9858211182246758e-05,
+      "loss": 1.6824,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9858110557987382e-05,
+      "loss": 1.797,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.985800989829043e-05,
+      "loss": 1.9285,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9857909203156268e-05,
+      "loss": 1.7193,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9857808472585248e-05,
+      "loss": 2.071,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.985770770657774e-05,
+      "loss": 1.9391,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.985760690513411e-05,
+      "loss": 1.959,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.985750606825471e-05,
+      "loss": 1.6609,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9857405195939904e-05,
+      "loss": 1.8168,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9857304288190063e-05,
+      "loss": 1.6298,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.985720334500554e-05,
+      "loss": 1.5715,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9857102366386706e-05,
+      "loss": 1.9339,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9857001352333914e-05,
+      "loss": 1.9379,
+      "step": 2834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.985690030284754e-05,
+      "loss": 2.0812,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9856799217927937e-05,
+      "loss": 2.0413,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9856698097575472e-05,
+      "loss": 2.3041,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9856596941790507e-05,
+      "loss": 1.9771,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.985649575057341e-05,
+      "loss": 2.1669,
+      "step": 2839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9856394523924544e-05,
+      "loss": 2.1472,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9856293261844266e-05,
+      "loss": 1.8587,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.9856191964332946e-05,
+      "loss": 1.7997,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9856090631390946e-05,
+      "loss": 2.0138,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9855989263018632e-05,
+      "loss": 1.8611,
+      "step": 2844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9855887859216366e-05,
+      "loss": 2.1201,
+      "step": 2845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9855786419984515e-05,
+      "loss": 2.1766,
+      "step": 2846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9855684945323443e-05,
+      "loss": 2.1677,
+      "step": 2847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9855583435233515e-05,
+      "loss": 1.6456,
+      "step": 2848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9855481889715095e-05,
+      "loss": 2.2235,
+      "step": 2849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9855380308768546e-05,
+      "loss": 1.7918,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.985527869239424e-05,
+      "loss": 2.0427,
+      "step": 2851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9855177040592536e-05,
+      "loss": 2.0735,
+      "step": 2852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.98550753533638e-05,
+      "loss": 1.8204,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9854973630708398e-05,
+      "loss": 1.6251,
+      "step": 2854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.98548718726267e-05,
+      "loss": 1.8821,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9854770079119064e-05,
+      "loss": 2.0588,
+      "step": 2856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9854668250185862e-05,
+      "loss": 1.7921,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.985456638582746e-05,
+      "loss": 1.7668,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9854464486044218e-05,
+      "loss": 2.0848,
+      "step": 2859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9854362550836514e-05,
+      "loss": 1.8673,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.98542605802047e-05,
+      "loss": 1.9539,
+      "step": 2861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9854158574149153e-05,
+      "loss": 1.7856,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9854056532670235e-05,
+      "loss": 1.9517,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9853954455768314e-05,
+      "loss": 1.7573,
+      "step": 2864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9853852343443757e-05,
+      "loss": 2.0343,
+      "step": 2865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9853750195696936e-05,
+      "loss": 1.8216,
+      "step": 2866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9853648012528207e-05,
+      "loss": 1.6959,
+      "step": 2867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9853545793937946e-05,
+      "loss": 1.6738,
+      "step": 2868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.985344353992652e-05,
+      "loss": 0.6934,
+      "step": 2869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.985334125049429e-05,
+      "loss": 2.2083,
+      "step": 2870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9853238925641632e-05,
+      "loss": 1.9491,
+      "step": 2871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9853136565368912e-05,
+      "loss": 1.1346,
+      "step": 2872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9853034169676492e-05,
+      "loss": 1.8799,
+      "step": 2873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9852931738564746e-05,
+      "loss": 2.2754,
+      "step": 2874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9852829272034043e-05,
+      "loss": 2.0022,
+      "step": 2875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9852726770084746e-05,
+      "loss": 1.9254,
+      "step": 2876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.985262423271723e-05,
+      "loss": 2.0075,
+      "step": 2877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9852521659931857e-05,
+      "loss": 1.8002,
+      "step": 2878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9852419051729e-05,
+      "loss": 1.8751,
+      "step": 2879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9852316408109022e-05,
+      "loss": 1.6517,
+      "step": 2880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.98522137290723e-05,
+      "loss": 1.9236,
+      "step": 2881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.98521110146192e-05,
+      "loss": 2.0039,
+      "step": 2882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9852008264750093e-05,
+      "loss": 1.9453,
+      "step": 2883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.985190547946534e-05,
+      "loss": 1.7672,
+      "step": 2884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9851802658765323e-05,
+      "loss": 1.9586,
+      "step": 2885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9851699802650402e-05,
+      "loss": 1.7726,
+      "step": 2886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.985159691112095e-05,
+      "loss": 1.9806,
+      "step": 2887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9851493984177338e-05,
+      "loss": 1.76,
+      "step": 2888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9851391021819934e-05,
+      "loss": 1.6561,
+      "step": 2889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.985128802404911e-05,
+      "loss": 1.9449,
+      "step": 2890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9851184990865234e-05,
+      "loss": 1.76,
+      "step": 2891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.985108192226868e-05,
+      "loss": 1.7879,
+      "step": 2892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9850978818259814e-05,
+      "loss": 2.3929,
+      "step": 2893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.985087567883901e-05,
+      "loss": 1.7381,
+      "step": 2894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.985077250400664e-05,
+      "loss": 1.8908,
+      "step": 2895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.985066929376307e-05,
+      "loss": 2.0847,
+      "step": 2896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.985056604810867e-05,
+      "loss": 1.8174,
+      "step": 2897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.985046276704382e-05,
+      "loss": 1.5903,
+      "step": 2898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9850359450568883e-05,
+      "loss": 2.132,
+      "step": 2899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9850256098684236e-05,
+      "loss": 1.8135,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9850152711390244e-05,
+      "loss": 2.2801,
+      "step": 2901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.985004928868729e-05,
+      "loss": 1.7828,
+      "step": 2902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9849945830575728e-05,
+      "loss": 1.6394,
+      "step": 2903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9849842337055944e-05,
+      "loss": 2.0347,
+      "step": 2904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.984973880812831e-05,
+      "loss": 1.8069,
+      "step": 2905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9849635243793192e-05,
+      "loss": 1.7663,
+      "step": 2906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9849531644050963e-05,
+      "loss": 2.2246,
+      "step": 2907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9849428008901998e-05,
+      "loss": 1.6372,
+      "step": 2908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9849324338346673e-05,
+      "loss": 1.9274,
+      "step": 2909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.984922063238535e-05,
+      "loss": 1.8739,
+      "step": 2910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.984911689101841e-05,
+      "loss": 2.0123,
+      "step": 2911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9849013114246227e-05,
+      "loss": 2.0498,
+      "step": 2912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9848909302069164e-05,
+      "loss": 1.6773,
+      "step": 2913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.984880545448761e-05,
+      "loss": 1.829,
+      "step": 2914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9848701571501926e-05,
+      "loss": 1.8991,
+      "step": 2915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9848597653112486e-05,
+      "loss": 1.5812,
+      "step": 2916
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.984849369931967e-05,
+      "loss": 1.8965,
+      "step": 2917
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.984838971012385e-05,
+      "loss": 1.8985,
+      "step": 2918
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9848285685525394e-05,
+      "loss": 1.8255,
+      "step": 2919
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.984818162552468e-05,
+      "loss": 1.8938,
+      "step": 2920
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 1.9471,
+      "step": 2921
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9847973399317975e-05,
+      "loss": 1.905,
+      "step": 2922
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9847869233112736e-05,
+      "loss": 1.7217,
+      "step": 2923
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9847765031506733e-05,
+      "loss": 2.072,
+      "step": 2924
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9847660794500344e-05,
+      "loss": 1.5742,
+      "step": 2925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.984755652209394e-05,
+      "loss": 2.0869,
+      "step": 2926
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9847452214287902e-05,
+      "loss": 1.6852,
+      "step": 2927
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9847347871082605e-05,
+      "loss": 1.6858,
+      "step": 2928
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.984724349247842e-05,
+      "loss": 1.6841,
+      "step": 2929
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.984713907847572e-05,
+      "loss": 2.1631,
+      "step": 2930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9847034629074888e-05,
+      "loss": 1.6917,
+      "step": 2931
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9846930144276294e-05,
+      "loss": 1.6925,
+      "step": 2932
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9846825624080312e-05,
+      "loss": 1.8885,
+      "step": 2933
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9846721068487325e-05,
+      "loss": 1.9542,
+      "step": 2934
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.98466164774977e-05,
+      "loss": 2.0129,
+      "step": 2935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.984651185111182e-05,
+      "loss": 1.9337,
+      "step": 2936
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.984640718933006e-05,
+      "loss": 1.6652,
+      "step": 2937
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9846302492152792e-05,
+      "loss": 1.8818,
+      "step": 2938
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9846197759580395e-05,
+      "loss": 2.2392,
+      "step": 2939
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.984609299161325e-05,
+      "loss": 1.8604,
+      "step": 2940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9845988188251724e-05,
+      "loss": 1.9336,
+      "step": 2941
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9845883349496202e-05,
+      "loss": 2.1099,
+      "step": 2942
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.984577847534706e-05,
+      "loss": 1.8345,
+      "step": 2943
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9845673565804668e-05,
+      "loss": 1.9124,
+      "step": 2944
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9845568620869405e-05,
+      "loss": 2.2111,
+      "step": 2945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.984546364054166e-05,
+      "loss": 1.9588,
+      "step": 2946
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.98453586248218e-05,
+      "loss": 1.959,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9845253573710197e-05,
+      "loss": 2.0708,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.984514848720724e-05,
+      "loss": 2.0297,
+      "step": 2949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9845043365313304e-05,
+      "loss": 1.7576,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9844938208028765e-05,
+      "loss": 2.0496,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9844833015354e-05,
+      "loss": 1.7614,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.984472778728939e-05,
+      "loss": 1.9629,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.984462252383531e-05,
+      "loss": 1.4475,
+      "step": 2954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.984451722499214e-05,
+      "loss": 2.1056,
+      "step": 2955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.984441189076026e-05,
+      "loss": 1.8388,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.984430652114005e-05,
+      "loss": 1.823,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.984420111613188e-05,
+      "loss": 1.7341,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9844095675736138e-05,
+      "loss": 1.9372,
+      "step": 2959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9843990199953204e-05,
+      "loss": 2.2788,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9843884688783446e-05,
+      "loss": 1.7576,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9843779142227258e-05,
+      "loss": 2.0746,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9843673560285007e-05,
+      "loss": 2.1919,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.984356794295708e-05,
+      "loss": 2.0038,
+      "step": 2964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9843462290243852e-05,
+      "loss": 2.1126,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9843356602145707e-05,
+      "loss": 1.9068,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9843250878663024e-05,
+      "loss": 1.9987,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9843145119796178e-05,
+      "loss": 1.7794,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9843039325545556e-05,
+      "loss": 1.5932,
+      "step": 2969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9842933495911532e-05,
+      "loss": 2.0736,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9842827630894496e-05,
+      "loss": 1.7168,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9842721730494817e-05,
+      "loss": 1.9848,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9842615794712884e-05,
+      "loss": 1.8973,
+      "step": 2973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.984250982354907e-05,
+      "loss": 3.5269,
+      "step": 2974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.984240381700377e-05,
+      "loss": 2.0491,
+      "step": 2975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.984229777507735e-05,
+      "loss": 1.7667,
+      "step": 2976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9842191697770197e-05,
+      "loss": 1.7566,
+      "step": 2977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9842085585082692e-05,
+      "loss": 1.9405,
+      "step": 2978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.984197943701522e-05,
+      "loss": 2.0431,
+      "step": 2979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.375,
+      "learning_rate": 1.984187325356816e-05,
+      "loss": 1.6903,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9841767034741886e-05,
+      "loss": 2.006,
+      "step": 2981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9841660780536792e-05,
+      "loss": 1.8683,
+      "step": 2982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9841554490953257e-05,
+      "loss": 1.6942,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.984144816599166e-05,
+      "loss": 1.7025,
+      "step": 2984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.984134180565238e-05,
+      "loss": 2.1504,
+      "step": 2985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.984123540993581e-05,
+      "loss": 1.7849,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.984112897884232e-05,
+      "loss": 1.8123,
+      "step": 2987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9841022512372302e-05,
+      "loss": 1.5745,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9840916010526138e-05,
+      "loss": 1.9957,
+      "step": 2989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9840809473304202e-05,
+      "loss": 2.1071,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.984070290070689e-05,
+      "loss": 1.9037,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9840596292734573e-05,
+      "loss": 2.008,
+      "step": 2992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9840489649387642e-05,
+      "loss": 1.7732,
+      "step": 2993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9840382970666477e-05,
+      "loss": 1.958,
+      "step": 2994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.984027625657146e-05,
+      "loss": 1.8658,
+      "step": 2995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.984016950710298e-05,
+      "loss": 1.8703,
+      "step": 2996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.984006272226142e-05,
+      "loss": 2.1141,
+      "step": 2997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9839955902047158e-05,
+      "loss": 1.7566,
+      "step": 2998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9839849046460584e-05,
+      "loss": 1.7611,
+      "step": 2999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.983974215550208e-05,
+      "loss": 2.0662,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9839635229172026e-05,
+      "loss": 1.6769,
+      "step": 3001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9839528267470814e-05,
+      "loss": 1.8905,
+      "step": 3002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.983942127039882e-05,
+      "loss": 1.8348,
+      "step": 3003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.983931423795644e-05,
+      "loss": 2.4784,
+      "step": 3004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.983920717014405e-05,
+      "loss": 1.6446,
+      "step": 3005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9839100066962035e-05,
+      "loss": 2.0918,
+      "step": 3006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9838992928410786e-05,
+      "loss": 1.8146,
+      "step": 3007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.983888575449068e-05,
+      "loss": 1.7119,
+      "step": 3008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9838778545202108e-05,
+      "loss": 1.7919,
+      "step": 3009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9838671300545454e-05,
+      "loss": 1.9113,
+      "step": 3010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9838564020521106e-05,
+      "loss": 1.6548,
+      "step": 3011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9838456705129446e-05,
+      "loss": 1.7611,
+      "step": 3012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.983834935437086e-05,
+      "loss": 1.9537,
+      "step": 3013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9838241968245735e-05,
+      "loss": 1.6347,
+      "step": 3014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9838134546754454e-05,
+      "loss": 2.1322,
+      "step": 3015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.983802708989741e-05,
+      "loss": 1.8611,
+      "step": 3016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9837919597674983e-05,
+      "loss": 1.7092,
+      "step": 3017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9837812070087563e-05,
+      "loss": 1.9575,
+      "step": 3018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9837704507135535e-05,
+      "loss": 1.6873,
+      "step": 3019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9837596908819285e-05,
+      "loss": 2.1628,
+      "step": 3020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.98374892751392e-05,
+      "loss": 2.1106,
+      "step": 3021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9837381606095668e-05,
+      "loss": 1.9439,
+      "step": 3022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9837273901689076e-05,
+      "loss": 2.0904,
+      "step": 3023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.983716616191981e-05,
+      "loss": 2.1354,
+      "step": 3024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9837058386788263e-05,
+      "loss": 1.8212,
+      "step": 3025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.983695057629481e-05,
+      "loss": 1.9751,
+      "step": 3026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9836842730439853e-05,
+      "loss": 2.2257,
+      "step": 3027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9836734849223766e-05,
+      "loss": 1.6384,
+      "step": 3028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.983662693264695e-05,
+      "loss": 1.8418,
+      "step": 3029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.983651898070978e-05,
+      "loss": 2.1364,
+      "step": 3030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9836410993412657e-05,
+      "loss": 2.0463,
+      "step": 3031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9836302970755957e-05,
+      "loss": 2.0114,
+      "step": 3032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.983619491274008e-05,
+      "loss": 1.7747,
+      "step": 3033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9836086819365402e-05,
+      "loss": 1.8175,
+      "step": 3034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9835978690632323e-05,
+      "loss": 1.7417,
+      "step": 3035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9835870526541222e-05,
+      "loss": 2.1193,
+      "step": 3036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9835762327092496e-05,
+      "loss": 2.1183,
+      "step": 3037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9835654092286528e-05,
+      "loss": 2.133,
+      "step": 3038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9835545822123714e-05,
+      "loss": 2.0166,
+      "step": 3039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9835437516604437e-05,
+      "loss": 1.7488,
+      "step": 3040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.9835329175729086e-05,
+      "loss": 1.9675,
+      "step": 3041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9835220799498055e-05,
+      "loss": 1.7747,
+      "step": 3042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.983511238791173e-05,
+      "loss": 1.8763,
+      "step": 3043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9835003940970504e-05,
+      "loss": 1.9102,
+      "step": 3044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9834895458674763e-05,
+      "loss": 1.7599,
+      "step": 3045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9834786941024897e-05,
+      "loss": 1.7585,
+      "step": 3046
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.98346783880213e-05,
+      "loss": 2.1087,
+      "step": 3047
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.983456979966436e-05,
+      "loss": 1.7857,
+      "step": 3048
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9834461175954467e-05,
+      "loss": 1.9267,
+      "step": 3049
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9834352516892013e-05,
+      "loss": 1.983,
+      "step": 3050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9834243822477382e-05,
+      "loss": 2.0221,
+      "step": 3051
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9834135092710974e-05,
+      "loss": 1.9406,
+      "step": 3052
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.983402632759318e-05,
+      "loss": 1.8544,
+      "step": 3053
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.983391752712438e-05,
+      "loss": 2.4032,
+      "step": 3054
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.983380869130498e-05,
+      "loss": 1.7966,
+      "step": 3055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.983369982013536e-05,
+      "loss": 1.676,
+      "step": 3056
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.983359091361591e-05,
+      "loss": 2.1475,
+      "step": 3057
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.983348197174703e-05,
+      "loss": 1.8085,
+      "step": 3058
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9833372994529107e-05,
+      "loss": 1.8957,
+      "step": 3059
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9833263981962535e-05,
+      "loss": 1.6627,
+      "step": 3060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9833154934047706e-05,
+      "loss": 1.9587,
+      "step": 3061
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.983304585078501e-05,
+      "loss": 1.9358,
+      "step": 3062
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9832936732174835e-05,
+      "loss": 1.8956,
+      "step": 3063
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9832827578217583e-05,
+      "loss": 2.2106,
+      "step": 3064
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.983271838891364e-05,
+      "loss": 1.9199,
+      "step": 3065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9832609164263397e-05,
+      "loss": 1.7366,
+      "step": 3066
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.983249990426725e-05,
+      "loss": 1.7371,
+      "step": 3067
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9832390608925595e-05,
+      "loss": 2.4179,
+      "step": 3068
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9832281278238818e-05,
+      "loss": 1.7296,
+      "step": 3069
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9832171912207315e-05,
+      "loss": 2.0094,
+      "step": 3070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.983206251083148e-05,
+      "loss": 1.8591,
+      "step": 3071
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9831953074111704e-05,
+      "loss": 1.6602,
+      "step": 3072
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9831843602048386e-05,
+      "loss": 2.0637,
+      "step": 3073
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.983173409464191e-05,
+      "loss": 1.6787,
+      "step": 3074
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.983162455189268e-05,
+      "loss": 1.898,
+      "step": 3075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9831514973801084e-05,
+      "loss": 1.9597,
+      "step": 3076
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9831405360367513e-05,
+      "loss": 1.9517,
+      "step": 3077
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9831295711592366e-05,
+      "loss": 1.9598,
+      "step": 3078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9831186027476037e-05,
+      "loss": 2.0162,
+      "step": 3079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.983107630801892e-05,
+      "loss": 2.4009,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.983096655322141e-05,
+      "loss": 1.7916,
+      "step": 3081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9830856763083898e-05,
+      "loss": 1.8833,
+      "step": 3082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.983074693760678e-05,
+      "loss": 1.9003,
+      "step": 3083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9830637076790453e-05,
+      "loss": 1.7754,
+      "step": 3084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.983052718063531e-05,
+      "loss": 2.2232,
+      "step": 3085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9830417249141747e-05,
+      "loss": 1.6291,
+      "step": 3086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.983030728231016e-05,
+      "loss": 2.04,
+      "step": 3087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9830197280140942e-05,
+      "loss": 2.0988,
+      "step": 3088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9830087242634488e-05,
+      "loss": 1.9514,
+      "step": 3089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9829977169791195e-05,
+      "loss": 1.9634,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.982986706161146e-05,
+      "loss": 1.974,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.982975691809568e-05,
+      "loss": 1.684,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9829646739244247e-05,
+      "loss": 1.8877,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9829536525057556e-05,
+      "loss": 1.7665,
+      "step": 3094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9829426275536007e-05,
+      "loss": 1.5044,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.982931599068e-05,
+      "loss": 1.8863,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.982920567048992e-05,
+      "loss": 1.7426,
+      "step": 3097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9829095314966175e-05,
+      "loss": 1.9476,
+      "step": 3098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9828984924109153e-05,
+      "loss": 2.1224,
+      "step": 3099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9828874497919256e-05,
+      "loss": 1.9767,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.982876403639688e-05,
+      "loss": 2.2533,
+      "step": 3101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9828653539542424e-05,
+      "loss": 1.9826,
+      "step": 3102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9828543007356277e-05,
+      "loss": 2.1158,
+      "step": 3103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9828432439838848e-05,
+      "loss": 2.2188,
+      "step": 3104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9828321836990527e-05,
+      "loss": 2.38,
+      "step": 3105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9828211198811712e-05,
+      "loss": 1.8097,
+      "step": 3106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9828100525302804e-05,
+      "loss": 1.8386,
+      "step": 3107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9827989816464196e-05,
+      "loss": 1.9494,
+      "step": 3108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.982787907229629e-05,
+      "loss": 1.5851,
+      "step": 3109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.982776829279949e-05,
+      "loss": 1.7266,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9827657477974176e-05,
+      "loss": 2.3755,
+      "step": 3111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9827546627820765e-05,
+      "loss": 2.0393,
+      "step": 3112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9827435742339646e-05,
+      "loss": 1.907,
+      "step": 3113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9827324821531216e-05,
+      "loss": 1.7677,
+      "step": 3114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9827213865395884e-05,
+      "loss": 2.0133,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9827102873934037e-05,
+      "loss": 1.974,
+      "step": 3116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.982699184714608e-05,
+      "loss": 1.6708,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9826880785032412e-05,
+      "loss": 2.0619,
+      "step": 3118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.982676968759343e-05,
+      "loss": 1.8812,
+      "step": 3119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9826658554829538e-05,
+      "loss": 1.8788,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.982654738674113e-05,
+      "loss": 2.026,
+      "step": 3121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.982643618332861e-05,
+      "loss": 1.8211,
+      "step": 3122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9826324944592375e-05,
+      "loss": 2.1371,
+      "step": 3123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.982621367053283e-05,
+      "loss": 1.8974,
+      "step": 3124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.9826102361150365e-05,
+      "loss": 1.633,
+      "step": 3125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9825991016445387e-05,
+      "loss": 2.091,
+      "step": 3126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9825879636418295e-05,
+      "loss": 1.7957,
+      "step": 3127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.982576822106949e-05,
+      "loss": 1.7577,
+      "step": 3128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.982565677039937e-05,
+      "loss": 2.1353,
+      "step": 3129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.982554528440834e-05,
+      "loss": 1.891,
+      "step": 3130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.98254337630968e-05,
+      "loss": 2.0018,
+      "step": 3131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9825322206465146e-05,
+      "loss": 1.5008,
+      "step": 3132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9825210614513782e-05,
+      "loss": 1.5237,
+      "step": 3133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.982509898724311e-05,
+      "loss": 1.7302,
+      "step": 3134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9824987324653534e-05,
+      "loss": 2.0921,
+      "step": 3135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9824875626745447e-05,
+      "loss": 1.8623,
+      "step": 3136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.982476389351926e-05,
+      "loss": 1.6075,
+      "step": 3137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.982465212497537e-05,
+      "loss": 1.7958,
+      "step": 3138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9824540321114177e-05,
+      "loss": 1.9808,
+      "step": 3139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9824428481936085e-05,
+      "loss": 1.6898,
+      "step": 3140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9824316607441495e-05,
+      "loss": 1.8561,
+      "step": 3141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9824204697630813e-05,
+      "loss": 1.7305,
+      "step": 3142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9824092752504437e-05,
+      "loss": 1.566,
+      "step": 3143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.982398077206277e-05,
+      "loss": 1.9427,
+      "step": 3144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.982386875630622e-05,
+      "loss": 2.0285,
+      "step": 3145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.982375670523518e-05,
+      "loss": 1.774,
+      "step": 3146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.982364461885006e-05,
+      "loss": 1.7293,
+      "step": 3147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.982353249715126e-05,
+      "loss": 1.9124,
+      "step": 3148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9823420340139184e-05,
+      "loss": 1.8202,
+      "step": 3149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9823308147814235e-05,
+      "loss": 1.9963,
+      "step": 3150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9823195920176817e-05,
+      "loss": 1.8839,
+      "step": 3151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9823083657227335e-05,
+      "loss": 1.9011,
+      "step": 3152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9822971358966187e-05,
+      "loss": 1.8195,
+      "step": 3153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9822859025393777e-05,
+      "loss": 1.9709,
+      "step": 3154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9822746656510516e-05,
+      "loss": 2.3292,
+      "step": 3155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9822634252316802e-05,
+      "loss": 1.7803,
+      "step": 3156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9822521812813046e-05,
+      "loss": 2.2253,
+      "step": 3157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.982240933799964e-05,
+      "loss": 1.954,
+      "step": 3158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9822296827877e-05,
+      "loss": 1.7774,
+      "step": 3159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9822184282445526e-05,
+      "loss": 1.9508,
+      "step": 3160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.982207170170562e-05,
+      "loss": 2.0942,
+      "step": 3161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.982195908565769e-05,
+      "loss": 1.7682,
+      "step": 3162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.982184643430214e-05,
+      "loss": 1.8186,
+      "step": 3163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9821733747639374e-05,
+      "loss": 1.8168,
+      "step": 3164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.98216210256698e-05,
+      "loss": 2.1482,
+      "step": 3165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.982150826839382e-05,
+      "loss": 1.4755,
+      "step": 3166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.982139547581184e-05,
+      "loss": 2.0599,
+      "step": 3167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9821282647924267e-05,
+      "loss": 2.3455,
+      "step": 3168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9821169784731503e-05,
+      "loss": 2.1084,
+      "step": 3169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.982105688623396e-05,
+      "loss": 2.2011,
+      "step": 3170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9820943952432038e-05,
+      "loss": 2.0476,
+      "step": 3171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9820830983326146e-05,
+      "loss": 1.9357,
+      "step": 3172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9820717978916687e-05,
+      "loss": 1.7251,
+      "step": 3173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.982060493920407e-05,
+      "loss": 1.9699,
+      "step": 3174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.98204918641887e-05,
+      "loss": 2.0272,
+      "step": 3175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.982037875387099e-05,
+      "loss": 2.1136,
+      "step": 3176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9820265608251335e-05,
+      "loss": 3.4386,
+      "step": 3177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.982015242733015e-05,
+      "loss": 2.0329,
+      "step": 3178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.982003921110784e-05,
+      "loss": 1.7538,
+      "step": 3179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.981992595958481e-05,
+      "loss": 1.6129,
+      "step": 3180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.981981267276147e-05,
+      "loss": 1.9287,
+      "step": 3181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9819699350638223e-05,
+      "loss": 1.7686,
+      "step": 3182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.981958599321548e-05,
+      "loss": 1.8581,
+      "step": 3183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.981947260049365e-05,
+      "loss": 2.0395,
+      "step": 3184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9819359172473133e-05,
+      "loss": 1.9841,
+      "step": 3185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9819245709154345e-05,
+      "loss": 1.5332,
+      "step": 3186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9819132210537692e-05,
+      "loss": 1.8436,
+      "step": 3187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.981901867662358e-05,
+      "loss": 2.1012,
+      "step": 3188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.981890510741242e-05,
+      "loss": 1.9524,
+      "step": 3189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9818791502904616e-05,
+      "loss": 1.9417,
+      "step": 3190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.981867786310058e-05,
+      "loss": 1.5019,
+      "step": 3191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9818564188000716e-05,
+      "loss": 2.2021,
+      "step": 3192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.625,
+      "learning_rate": 1.981845047760544e-05,
+      "loss": 2.0257,
+      "step": 3193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9818336731915154e-05,
+      "loss": 1.9201,
+      "step": 3194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.981822295093027e-05,
+      "loss": 1.8612,
+      "step": 3195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9818109134651195e-05,
+      "loss": 1.8087,
+      "step": 3196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9817995283078342e-05,
+      "loss": 1.8862,
+      "step": 3197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9817881396212116e-05,
+      "loss": 1.8864,
+      "step": 3198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9817767474052927e-05,
+      "loss": 2.1656,
+      "step": 3199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9817653516601186e-05,
+      "loss": 2.0541,
+      "step": 3200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.981753952385731e-05,
+      "loss": 2.0536,
+      "step": 3201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9817425495821692e-05,
+      "loss": 1.6714,
+      "step": 3202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9817311432494756e-05,
+      "loss": 1.8746,
+      "step": 3203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9817197333876904e-05,
+      "loss": 1.7229,
+      "step": 3204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9817083199968552e-05,
+      "loss": 1.7236,
+      "step": 3205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9816969030770107e-05,
+      "loss": 1.671,
+      "step": 3206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9816854826281976e-05,
+      "loss": 1.9079,
+      "step": 3207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.981674058650458e-05,
+      "loss": 2.0557,
+      "step": 3208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9816626311438318e-05,
+      "loss": 1.7817,
+      "step": 3209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.981651200108361e-05,
+      "loss": 1.9219,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.981639765544086e-05,
+      "loss": 2.0649,
+      "step": 3211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9816283274510485e-05,
+      "loss": 1.718,
+      "step": 3212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.981616885829289e-05,
+      "loss": 1.7958,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9816054406788487e-05,
+      "loss": 2.1738,
+      "step": 3214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9815939919997695e-05,
+      "loss": 1.8929,
+      "step": 3215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.981582539792092e-05,
+      "loss": 1.909,
+      "step": 3216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9815710840558574e-05,
+      "loss": 1.6978,
+      "step": 3217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9815596247911063e-05,
+      "loss": 1.8982,
+      "step": 3218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.981548161997881e-05,
+      "loss": 1.9864,
+      "step": 3219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9815366956762224e-05,
+      "loss": 1.8712,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9815252258261713e-05,
+      "loss": 2.149,
+      "step": 3221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.981513752447769e-05,
+      "loss": 1.9468,
+      "step": 3222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.981502275541057e-05,
+      "loss": 1.9113,
+      "step": 3223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.981490795106076e-05,
+      "loss": 1.9125,
+      "step": 3224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.981479311142868e-05,
+      "loss": 2.0609,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9814678236514745e-05,
+      "loss": 1.771,
+      "step": 3226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.981456332631936e-05,
+      "loss": 1.7696,
+      "step": 3227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.981444838084294e-05,
+      "loss": 1.9747,
+      "step": 3228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9814333400085897e-05,
+      "loss": 2.081,
+      "step": 3229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9814218384048648e-05,
+      "loss": 1.7511,
+      "step": 3230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9814103332731603e-05,
+      "loss": 1.9386,
+      "step": 3231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.981398824613518e-05,
+      "loss": 1.8683,
+      "step": 3232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9813873124259794e-05,
+      "loss": 1.932,
+      "step": 3233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9813757967105847e-05,
+      "loss": 2.0054,
+      "step": 3234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9813642774673766e-05,
+      "loss": 2.0008,
+      "step": 3235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.981352754696396e-05,
+      "loss": 1.9041,
+      "step": 3236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.981341228397684e-05,
+      "loss": 1.7204,
+      "step": 3237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9813296985712825e-05,
+      "loss": 1.8692,
+      "step": 3238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.981318165217233e-05,
+      "loss": 1.9846,
+      "step": 3239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9813066283355766e-05,
+      "loss": 1.5237,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9812950879263552e-05,
+      "loss": 1.9232,
+      "step": 3241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.98128354398961e-05,
+      "loss": 2.0045,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9812719965253823e-05,
+      "loss": 1.9278,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9812604455337138e-05,
+      "loss": 1.806,
+      "step": 3244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.981248891014646e-05,
+      "loss": 1.9953,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9812373329682207e-05,
+      "loss": 1.9153,
+      "step": 3246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9812257713944795e-05,
+      "loss": 2.0828,
+      "step": 3247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.981214206293463e-05,
+      "loss": 1.8754,
+      "step": 3248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.981202637665214e-05,
+      "loss": 2.1879,
+      "step": 3249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9811910655097738e-05,
+      "loss": 1.8313,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.981179489827183e-05,
+      "loss": 2.0018,
+      "step": 3251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9811679106174846e-05,
+      "loss": 1.9499,
+      "step": 3252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9811563278807194e-05,
+      "loss": 0.761,
+      "step": 3253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9811447416169293e-05,
+      "loss": 2.0356,
+      "step": 3254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9811331518261557e-05,
+      "loss": 1.9031,
+      "step": 3255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9811215585084405e-05,
+      "loss": 1.7042,
+      "step": 3256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9811099616638256e-05,
+      "loss": 1.8189,
+      "step": 3257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.981098361292352e-05,
+      "loss": 1.8664,
+      "step": 3258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.981086757394062e-05,
+      "loss": 2.0816,
+      "step": 3259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9810751499689973e-05,
+      "loss": 1.9732,
+      "step": 3260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9810635390171994e-05,
+      "loss": 2.4235,
+      "step": 3261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.98105192453871e-05,
+      "loss": 1.8055,
+      "step": 3262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9810403065335708e-05,
+      "loss": 2.1234,
+      "step": 3263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9810286850018238e-05,
+      "loss": 1.5466,
+      "step": 3264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9810170599435107e-05,
+      "loss": 1.8355,
+      "step": 3265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9810054313586735e-05,
+      "loss": 2.0687,
+      "step": 3266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9809937992473535e-05,
+      "loss": 1.7564,
+      "step": 3267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9809821636095928e-05,
+      "loss": 1.7274,
+      "step": 3268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9809705244454333e-05,
+      "loss": 1.9742,
+      "step": 3269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.980958881754917e-05,
+      "loss": 1.9267,
+      "step": 3270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9809472355380853e-05,
+      "loss": 1.6288,
+      "step": 3271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.98093558579498e-05,
+      "loss": 1.6079,
+      "step": 3272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9809239325256434e-05,
+      "loss": 2.3135,
+      "step": 3273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9809122757301177e-05,
+      "loss": 1.9411,
+      "step": 3274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9809006154084437e-05,
+      "loss": 2.094,
+      "step": 3275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9808889515606644e-05,
+      "loss": 2.0968,
+      "step": 3276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.980877284186821e-05,
+      "loss": 1.6223,
+      "step": 3277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9808656132869558e-05,
+      "loss": 2.0612,
+      "step": 3278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.980853938861111e-05,
+      "loss": 1.9088,
+      "step": 3279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.980842260909328e-05,
+      "loss": 2.0572,
+      "step": 3280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9808305794316488e-05,
+      "loss": 2.0534,
+      "step": 3281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.980818894428116e-05,
+      "loss": 1.7778,
+      "step": 3282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.980807205898771e-05,
+      "loss": 2.2595,
+      "step": 3283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9807955138436562e-05,
+      "loss": 2.081,
+      "step": 3284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9807838182628134e-05,
+      "loss": 2.0548,
+      "step": 3285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.980772119156285e-05,
+      "loss": 1.9374,
+      "step": 3286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9807604165241123e-05,
+      "loss": 1.9188,
+      "step": 3287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.980748710366338e-05,
+      "loss": 1.8583,
+      "step": 3288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.980737000683004e-05,
+      "loss": 1.7973,
+      "step": 3289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9807252874741527e-05,
+      "loss": 2.1236,
+      "step": 3290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.980713570739826e-05,
+      "loss": 2.1409,
+      "step": 3291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9807018504800653e-05,
+      "loss": 1.5381,
+      "step": 3292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.980690126694914e-05,
+      "loss": 1.8217,
+      "step": 3293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9806783993844135e-05,
+      "loss": 1.9447,
+      "step": 3294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.980666668548606e-05,
+      "loss": 2.1748,
+      "step": 3295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9806549341875337e-05,
+      "loss": 1.8033,
+      "step": 3296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9806431963012387e-05,
+      "loss": 2.1707,
+      "step": 3297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9806314548897637e-05,
+      "loss": 1.9143,
+      "step": 3298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.98061970995315e-05,
+      "loss": 1.7204,
+      "step": 3299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.980607961491441e-05,
+      "loss": 2.0471,
+      "step": 3300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.980596209504678e-05,
+      "loss": 2.1682,
+      "step": 3301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9805844539929033e-05,
+      "loss": 1.882,
+      "step": 3302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9805726949561593e-05,
+      "loss": 1.9092,
+      "step": 3303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.980560932394489e-05,
+      "loss": 1.9346,
+      "step": 3304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9805491663079337e-05,
+      "loss": 1.9552,
+      "step": 3305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9805373966965357e-05,
+      "loss": 1.76,
+      "step": 3306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.980525623560338e-05,
+      "loss": 1.8816,
+      "step": 3307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9805138468993827e-05,
+      "loss": 1.9351,
+      "step": 3308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9805020667137117e-05,
+      "loss": 2.1083,
+      "step": 3309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.980490283003368e-05,
+      "loss": 1.6059,
+      "step": 3310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9804784957683934e-05,
+      "loss": 2.0437,
+      "step": 3311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.9804667050088304e-05,
+      "loss": 2.2418,
+      "step": 3312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9804549107247216e-05,
+      "loss": 1.7944,
+      "step": 3313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9804431129161094e-05,
+      "loss": 1.9865,
+      "step": 3314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.9804313115830358e-05,
+      "loss": 2.2182,
+      "step": 3315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9804195067255435e-05,
+      "loss": 2.3637,
+      "step": 3316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.980407698343675e-05,
+      "loss": 2.0746,
+      "step": 3317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9803958864374728e-05,
+      "loss": 1.8042,
+      "step": 3318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.980384071006979e-05,
+      "loss": 2.002,
+      "step": 3319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9803722520522365e-05,
+      "loss": 2.1055,
+      "step": 3320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9803604295732875e-05,
+      "loss": 1.7703,
+      "step": 3321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.980348603570175e-05,
+      "loss": 1.7339,
+      "step": 3322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9803367740429406e-05,
+      "loss": 2.0433,
+      "step": 3323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9803249409916277e-05,
+      "loss": 2.0067,
+      "step": 3324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.980313104416278e-05,
+      "loss": 1.9066,
+      "step": 3325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9803012643169352e-05,
+      "loss": 1.8479,
+      "step": 3326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.980289420693641e-05,
+      "loss": 1.8399,
+      "step": 3327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9802775735464377e-05,
+      "loss": 1.7207,
+      "step": 3328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9802657228753687e-05,
+      "loss": 2.0116,
+      "step": 3329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9802538686804766e-05,
+      "loss": 1.6466,
+      "step": 3330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.980242010961803e-05,
+      "loss": 2.0319,
+      "step": 3331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9802301497193915e-05,
+      "loss": 1.9869,
+      "step": 3332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9802182849532845e-05,
+      "loss": 1.803,
+      "step": 3333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9802064166635243e-05,
+      "loss": 1.6832,
+      "step": 3334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.980194544850154e-05,
+      "loss": 1.8571,
+      "step": 3335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9801826695132162e-05,
+      "loss": 1.7886,
+      "step": 3336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9801707906527535e-05,
+      "loss": 2.1278,
+      "step": 3337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9801589082688085e-05,
+      "loss": 1.8809,
+      "step": 3338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.980147022361424e-05,
+      "loss": 1.8306,
+      "step": 3339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.980135132930643e-05,
+      "loss": 1.9836,
+      "step": 3340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.980123239976508e-05,
+      "loss": 1.8968,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9801113434990614e-05,
+      "loss": 1.858,
+      "step": 3342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9800994434983466e-05,
+      "loss": 1.7303,
+      "step": 3343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.980087539974406e-05,
+      "loss": 1.9754,
+      "step": 3344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9800756329272825e-05,
+      "loss": 1.7889,
+      "step": 3345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.980063722357019e-05,
+      "loss": 2.0561,
+      "step": 3346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.980051808263658e-05,
+      "loss": 1.8369,
+      "step": 3347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9800398906472426e-05,
+      "loss": 1.6988,
+      "step": 3348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9800279695078154e-05,
+      "loss": 1.8181,
+      "step": 3349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9800160448454195e-05,
+      "loss": 2.2257,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9800041166600975e-05,
+      "loss": 1.7161,
+      "step": 3351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.979992184951893e-05,
+      "loss": 1.8208,
+      "step": 3352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.979980249720848e-05,
+      "loss": 1.7912,
+      "step": 3353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9799683109670055e-05,
+      "loss": 1.9273,
+      "step": 3354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.979956368690409e-05,
+      "loss": 1.6648,
+      "step": 3355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9799444228911012e-05,
+      "loss": 2.0057,
+      "step": 3356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.979932473569125e-05,
+      "loss": 1.8446,
+      "step": 3357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.979920520724523e-05,
+      "loss": 2.0535,
+      "step": 3358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9799085643573388e-05,
+      "loss": 2.0015,
+      "step": 3359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9798966044676148e-05,
+      "loss": 2.0604,
+      "step": 3360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9798846410553946e-05,
+      "loss": 1.9495,
+      "step": 3361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9798726741207204e-05,
+      "loss": 2.0041,
+      "step": 3362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.979860703663636e-05,
+      "loss": 1.8078,
+      "step": 3363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.979848729684184e-05,
+      "loss": 1.9029,
+      "step": 3364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9798367521824073e-05,
+      "loss": 1.8066,
+      "step": 3365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9798247711583493e-05,
+      "loss": 1.821,
+      "step": 3366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9798127866120533e-05,
+      "loss": 1.7648,
+      "step": 3367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9798007985435615e-05,
+      "loss": 1.7704,
+      "step": 3368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.979788806952918e-05,
+      "loss": 2.0267,
+      "step": 3369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.979776811840165e-05,
+      "loss": 2.1577,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9797648132053466e-05,
+      "loss": 1.7999,
+      "step": 3371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9797528110485055e-05,
+      "loss": 1.9526,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9797408053696844e-05,
+      "loss": 1.6635,
+      "step": 3373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9797287961689267e-05,
+      "loss": 1.9971,
+      "step": 3374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.979716783446276e-05,
+      "loss": 2.3407,
+      "step": 3375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.979704767201775e-05,
+      "loss": 1.9773,
+      "step": 3376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.979692747435467e-05,
+      "loss": 1.4662,
+      "step": 3377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.979680724147395e-05,
+      "loss": 2.0633,
+      "step": 3378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.979668697337603e-05,
+      "loss": 1.8768,
+      "step": 3379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9796566670061334e-05,
+      "loss": 2.0288,
+      "step": 3380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.97964463315303e-05,
+      "loss": 1.7026,
+      "step": 3381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9796325957783356e-05,
+      "loss": 1.9956,
+      "step": 3382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.979620554882094e-05,
+      "loss": 1.8158,
+      "step": 3383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9796085104643478e-05,
+      "loss": 1.6523,
+      "step": 3384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.979596462525141e-05,
+      "loss": 1.3774,
+      "step": 3385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9795844110645164e-05,
+      "loss": 1.8063,
+      "step": 3386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9795723560825174e-05,
+      "loss": 1.7502,
+      "step": 3387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9795602975791875e-05,
+      "loss": 2.0987,
+      "step": 3388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.97954823555457e-05,
+      "loss": 1.8101,
+      "step": 3389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9795361700087087e-05,
+      "loss": 1.7654,
+      "step": 3390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.979524100941646e-05,
+      "loss": 2.1954,
+      "step": 3391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9795120283534258e-05,
+      "loss": 1.8403,
+      "step": 3392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.979499952244092e-05,
+      "loss": 1.6628,
+      "step": 3393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.979487872613687e-05,
+      "loss": 1.7687,
+      "step": 3394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.979475789462255e-05,
+      "loss": 1.9299,
+      "step": 3395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.979463702789839e-05,
+      "loss": 2.047,
+      "step": 3396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9794516125964826e-05,
+      "loss": 1.9931,
+      "step": 3397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9794395188822297e-05,
+      "loss": 1.869,
+      "step": 3398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.979427421647123e-05,
+      "loss": 2.0397,
+      "step": 3399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9794153208912063e-05,
+      "loss": 2.1743,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9794032166145233e-05,
+      "loss": 1.9609,
+      "step": 3401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.9793911088171173e-05,
+      "loss": 2.2254,
+      "step": 3402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9793789974990318e-05,
+      "loss": 2.0042,
+      "step": 3403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9793668826603105e-05,
+      "loss": 2.1703,
+      "step": 3404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.979354764300997e-05,
+      "loss": 1.705,
+      "step": 3405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9793426424211346e-05,
+      "loss": 1.9099,
+      "step": 3406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9793305170207672e-05,
+      "loss": 1.749,
+      "step": 3407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9793183880999377e-05,
+      "loss": 1.6476,
+      "step": 3408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9793062556586906e-05,
+      "loss": 1.8619,
+      "step": 3409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9792941196970692e-05,
+      "loss": 1.9412,
+      "step": 3410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9792819802151165e-05,
+      "loss": 1.9022,
+      "step": 3411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9792698372128773e-05,
+      "loss": 1.9555,
+      "step": 3412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9792576906903945e-05,
+      "loss": 2.2374,
+      "step": 3413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.979245540647712e-05,
+      "loss": 1.6802,
+      "step": 3414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.979233387084873e-05,
+      "loss": 1.8254,
+      "step": 3415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9792212300019215e-05,
+      "loss": 1.8611,
+      "step": 3416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9792090693989014e-05,
+      "loss": 1.7329,
+      "step": 3417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9791969052758563e-05,
+      "loss": 2.0424,
+      "step": 3418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.97918473763283e-05,
+      "loss": 2.0207,
+      "step": 3419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.979172566469866e-05,
+      "loss": 1.6794,
+      "step": 3420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9791603917870082e-05,
+      "loss": 1.6902,
+      "step": 3421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9791482135843006e-05,
+      "loss": 1.9768,
+      "step": 3422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9791360318617862e-05,
+      "loss": 2.0279,
+      "step": 3423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9791238466195098e-05,
+      "loss": 1.9714,
+      "step": 3424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9791116578575146e-05,
+      "loss": 2.6382,
+      "step": 3425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9790994655758445e-05,
+      "loss": 1.8103,
+      "step": 3426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9790872697745437e-05,
+      "loss": 1.883,
+      "step": 3427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.979075070453655e-05,
+      "loss": 1.9205,
+      "step": 3428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9790628676132236e-05,
+      "loss": 1.8408,
+      "step": 3429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9790506612532922e-05,
+      "loss": 1.8195,
+      "step": 3430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9790384513739054e-05,
+      "loss": 2.059,
+      "step": 3431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9790262379751076e-05,
+      "loss": 2.0988,
+      "step": 3432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.979014021056941e-05,
+      "loss": 1.9261,
+      "step": 3433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9790018006194512e-05,
+      "loss": 1.7922,
+      "step": 3434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.978989576662681e-05,
+      "loss": 1.8406,
+      "step": 3435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.978977349186675e-05,
+      "loss": 2.0578,
+      "step": 3436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9789651181914773e-05,
+      "loss": 1.896,
+      "step": 3437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.978952883677131e-05,
+      "loss": 2.2344,
+      "step": 3438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.978940645643681e-05,
+      "loss": 2.1067,
+      "step": 3439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9789284040911705e-05,
+      "loss": 1.9505,
+      "step": 3440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.978916159019644e-05,
+      "loss": 1.8897,
+      "step": 3441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9789039104291453e-05,
+      "loss": 1.7828,
+      "step": 3442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.978891658319719e-05,
+      "loss": 1.8337,
+      "step": 3443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9788794026914085e-05,
+      "loss": 1.724,
+      "step": 3444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9788671435442575e-05,
+      "loss": 1.7484,
+      "step": 3445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.9788548808783115e-05,
+      "loss": 1.9918,
+      "step": 3446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9788426146936133e-05,
+      "loss": 2.0148,
+      "step": 3447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.978830344990207e-05,
+      "loss": 1.7671,
+      "step": 3448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9788180717681377e-05,
+      "loss": 1.8373,
+      "step": 3449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.978805795027449e-05,
+      "loss": 2.1151,
+      "step": 3450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9787935147681844e-05,
+      "loss": 1.83,
+      "step": 3451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.978781230990389e-05,
+      "loss": 1.8406,
+      "step": 3452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9787689436941062e-05,
+      "loss": 2.3313,
+      "step": 3453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9787566528793806e-05,
+      "loss": 1.837,
+      "step": 3454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9787443585462564e-05,
+      "loss": 1.6565,
+      "step": 3455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9787320606947778e-05,
+      "loss": 1.9264,
+      "step": 3456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9787197593249887e-05,
+      "loss": 2.194,
+      "step": 3457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9787074544369337e-05,
+      "loss": 1.9183,
+      "step": 3458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.978695146030657e-05,
+      "loss": 1.9317,
+      "step": 3459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9786828341062023e-05,
+      "loss": 1.9686,
+      "step": 3460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9786705186636143e-05,
+      "loss": 1.9098,
+      "step": 3461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9786581997029373e-05,
+      "loss": 1.9916,
+      "step": 3462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9786458772242157e-05,
+      "loss": 1.8122,
+      "step": 3463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9786335512274935e-05,
+      "loss": 2.4544,
+      "step": 3464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.978621221712815e-05,
+      "loss": 1.7808,
+      "step": 3465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.978608888680225e-05,
+      "loss": 1.8272,
+      "step": 3466
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.978596552129767e-05,
+      "loss": 1.7941,
+      "step": 3467
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9785842120614864e-05,
+      "loss": 2.1891,
+      "step": 3468
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9785718684754264e-05,
+      "loss": 1.9826,
+      "step": 3469
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9785595213716325e-05,
+      "loss": 1.7576,
+      "step": 3470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9785471707501484e-05,
+      "loss": 2.0633,
+      "step": 3471
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9785348166110184e-05,
+      "loss": 2.0704,
+      "step": 3472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.978522458954287e-05,
+      "loss": 1.8697,
+      "step": 3473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9785100977799992e-05,
+      "loss": 1.9823,
+      "step": 3474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9784977330881987e-05,
+      "loss": 1.8511,
+      "step": 3475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9784853648789303e-05,
+      "loss": 1.8473,
+      "step": 3476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.978472993152239e-05,
+      "loss": 1.6893,
+      "step": 3477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.978460617908168e-05,
+      "loss": 2.0807,
+      "step": 3478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9784482391467628e-05,
+      "loss": 1.611,
+      "step": 3479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9784358568680673e-05,
+      "loss": 1.708,
+      "step": 3480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9784234710721266e-05,
+      "loss": 1.7195,
+      "step": 3481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9784110817589845e-05,
+      "loss": 1.9181,
+      "step": 3482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9783986889286865e-05,
+      "loss": 1.8393,
+      "step": 3483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9783862925812763e-05,
+      "loss": 1.7535,
+      "step": 3484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.978373892716799e-05,
+      "loss": 1.8602,
+      "step": 3485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.978361489335298e-05,
+      "loss": 1.8089,
+      "step": 3486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.97834908243682e-05,
+      "loss": 1.651,
+      "step": 3487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9783366720214078e-05,
+      "loss": 1.8872,
+      "step": 3488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9783242580891067e-05,
+      "loss": 1.7627,
+      "step": 3489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9783118406399614e-05,
+      "loss": 1.9614,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9782994196740164e-05,
+      "loss": 2.0602,
+      "step": 3491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9782869951913165e-05,
+      "loss": 2.0071,
+      "step": 3492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5,
+      "learning_rate": 1.978274567191906e-05,
+      "loss": 2.3284,
+      "step": 3493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9782621356758297e-05,
+      "loss": 1.6622,
+      "step": 3494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9782497006431326e-05,
+      "loss": 1.9771,
+      "step": 3495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.978237262093859e-05,
+      "loss": 2.0985,
+      "step": 3496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9782248200280537e-05,
+      "loss": 1.7402,
+      "step": 3497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9782123744457618e-05,
+      "loss": 1.9208,
+      "step": 3498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9781999253470275e-05,
+      "loss": 1.6304,
+      "step": 3499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.978187472731896e-05,
+      "loss": 1.9647,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9781750166004117e-05,
+      "loss": 1.8464,
+      "step": 3501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9781625569526196e-05,
+      "loss": 1.6627,
+      "step": 3502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9781500937885644e-05,
+      "loss": 2.5002,
+      "step": 3503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.978137627108291e-05,
+      "loss": 2.0217,
+      "step": 3504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.978125156911844e-05,
+      "loss": 2.1956,
+      "step": 3505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.978112683199269e-05,
+      "loss": 1.8021,
+      "step": 3506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9781002059706097e-05,
+      "loss": 1.9753,
+      "step": 3507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9780877252259113e-05,
+      "loss": 2.0392,
+      "step": 3508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9780752409652192e-05,
+      "loss": 1.7153,
+      "step": 3509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9780627531885776e-05,
+      "loss": 1.8843,
+      "step": 3510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.978050261896032e-05,
+      "loss": 1.8843,
+      "step": 3511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.978037767087627e-05,
+      "loss": 1.923,
+      "step": 3512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9780252687634075e-05,
+      "loss": 1.467,
+      "step": 3513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9780127669234183e-05,
+      "loss": 2.1009,
+      "step": 3514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9780002615677044e-05,
+      "loss": 1.9776,
+      "step": 3515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.977987752696311e-05,
+      "loss": 1.8984,
+      "step": 3516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.977975240309283e-05,
+      "loss": 1.7226,
+      "step": 3517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9779627244066652e-05,
+      "loss": 1.7781,
+      "step": 3518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9779502049885028e-05,
+      "loss": 1.7957,
+      "step": 3519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9779376820548406e-05,
+      "loss": 2.1481,
+      "step": 3520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.977925155605724e-05,
+      "loss": 1.9897,
+      "step": 3521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9779126256411974e-05,
+      "loss": 1.9075,
+      "step": 3522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.977900092161306e-05,
+      "loss": 2.2505,
+      "step": 3523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9778875551660954e-05,
+      "loss": 1.7673,
+      "step": 3524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.97787501465561e-05,
+      "loss": 1.9728,
+      "step": 3525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9778624706298958e-05,
+      "loss": 2.1025,
+      "step": 3526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9778499230889968e-05,
+      "loss": 1.7973,
+      "step": 3527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9778373720329585e-05,
+      "loss": 1.6479,
+      "step": 3528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9778248174618262e-05,
+      "loss": 2.6302,
+      "step": 3529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9778122593756452e-05,
+      "loss": 1.9418,
+      "step": 3530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9777996977744597e-05,
+      "loss": 1.897,
+      "step": 3531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0,
+      "learning_rate": 1.977787132658316e-05,
+      "loss": 2.2931,
+      "step": 3532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.977774564027259e-05,
+      "loss": 1.8718,
+      "step": 3533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9777619918813332e-05,
+      "loss": 1.569,
+      "step": 3534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9777494162205846e-05,
+      "loss": 2.0646,
+      "step": 3535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9777368370450582e-05,
+      "loss": 1.9458,
+      "step": 3536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9777242543547987e-05,
+      "loss": 2.1874,
+      "step": 3537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.977711668149852e-05,
+      "loss": 2.7772,
+      "step": 3538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9776990784302632e-05,
+      "loss": 1.6828,
+      "step": 3539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.977686485196077e-05,
+      "loss": 1.899,
+      "step": 3540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9776738884473393e-05,
+      "loss": 1.6409,
+      "step": 3541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9776612881840952e-05,
+      "loss": 1.9727,
+      "step": 3542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.97764868440639e-05,
+      "loss": 1.8559,
+      "step": 3543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.977636077114269e-05,
+      "loss": 2.1123,
+      "step": 3544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9776234663077776e-05,
+      "loss": 1.689,
+      "step": 3545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.977610851986961e-05,
+      "loss": 2.0027,
+      "step": 3546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9775982341518645e-05,
+      "loss": 1.8754,
+      "step": 3547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9775856128025335e-05,
+      "loss": 2.2453,
+      "step": 3548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9775729879390135e-05,
+      "loss": 1.6884,
+      "step": 3549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9775603595613496e-05,
+      "loss": 2.1364,
+      "step": 3550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9775477276695876e-05,
+      "loss": 1.7485,
+      "step": 3551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.977535092263773e-05,
+      "loss": 1.8911,
+      "step": 3552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9775224533439506e-05,
+      "loss": 1.9173,
+      "step": 3553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.977509810910166e-05,
+      "loss": 1.9537,
+      "step": 3554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9774971649624648e-05,
+      "loss": 1.8807,
+      "step": 3555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.977484515500893e-05,
+      "loss": 1.7608,
+      "step": 3556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.977471862525495e-05,
+      "loss": 1.7473,
+      "step": 3557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.977459206036317e-05,
+      "loss": 1.8136,
+      "step": 3558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.977446546033404e-05,
+      "loss": 2.0181,
+      "step": 3559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9774338825168024e-05,
+      "loss": 1.9884,
+      "step": 3560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.977421215486557e-05,
+      "loss": 1.7241,
+      "step": 3561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.977408544942713e-05,
+      "loss": 2.1876,
+      "step": 3562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.977395870885317e-05,
+      "loss": 2.026,
+      "step": 3563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9773831933144138e-05,
+      "loss": 1.5958,
+      "step": 3564
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.977370512230049e-05,
+      "loss": 1.808,
+      "step": 3565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9773578276322686e-05,
+      "loss": 1.7518,
+      "step": 3566
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9773451395211176e-05,
+      "loss": 2.1368,
+      "step": 3567
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9773324478966423e-05,
+      "loss": 1.8531,
+      "step": 3568
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.977319752758888e-05,
+      "loss": 1.7748,
+      "step": 3569
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9773070541079002e-05,
+      "loss": 1.8749,
+      "step": 3570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9772943519437245e-05,
+      "loss": 1.9932,
+      "step": 3571
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.977281646266407e-05,
+      "loss": 1.8193,
+      "step": 3572
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.977268937075993e-05,
+      "loss": 1.9089,
+      "step": 3573
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9772562243725284e-05,
+      "loss": 1.9393,
+      "step": 3574
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9772435081560585e-05,
+      "loss": 1.8579,
+      "step": 3575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9772307884266296e-05,
+      "loss": 1.715,
+      "step": 3576
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.977218065184287e-05,
+      "loss": 1.7179,
+      "step": 3577
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9772053384290766e-05,
+      "loss": 2.0258,
+      "step": 3578
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.977192608161044e-05,
+      "loss": 1.7304,
+      "step": 3579
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9771798743802352e-05,
+      "loss": 2.8961,
+      "step": 3580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9771671370866955e-05,
+      "loss": 1.682,
+      "step": 3581
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9771543962804712e-05,
+      "loss": 1.9967,
+      "step": 3582
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9771416519616084e-05,
+      "loss": 2.0293,
+      "step": 3583
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.977128904130152e-05,
+      "loss": 1.8318,
+      "step": 3584
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9771161527861485e-05,
+      "loss": 2.0479,
+      "step": 3585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.977103397929643e-05,
+      "loss": 1.5839,
+      "step": 3586
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9770906395606827e-05,
+      "loss": 2.1944,
+      "step": 3587
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.977077877679312e-05,
+      "loss": 2.1511,
+      "step": 3588
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9770651122855776e-05,
+      "loss": 1.8806,
+      "step": 3589
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9770523433795252e-05,
+      "loss": 1.8648,
+      "step": 3590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.977039570961201e-05,
+      "loss": 1.6987,
+      "step": 3591
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.97702679503065e-05,
+      "loss": 1.8896,
+      "step": 3592
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.977014015587919e-05,
+      "loss": 1.7031,
+      "step": 3593
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9770012326330537e-05,
+      "loss": 1.8617,
+      "step": 3594
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9769884461661e-05,
+      "loss": 1.9826,
+      "step": 3595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.976975656187104e-05,
+      "loss": 1.7524,
+      "step": 3596
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9769628626961114e-05,
+      "loss": 1.837,
+      "step": 3597
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.9769500656931684e-05,
+      "loss": 1.8813,
+      "step": 3598
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9769372651783207e-05,
+      "loss": 1.6116,
+      "step": 3599
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.976924461151615e-05,
+      "loss": 1.8997,
+      "step": 3600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9769116536130965e-05,
+      "loss": 1.7551,
+      "step": 3601
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.976898842562812e-05,
+      "loss": 1.6801,
+      "step": 3602
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.976886028000807e-05,
+      "loss": 1.9291,
+      "step": 3603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.976873209927128e-05,
+      "loss": 1.768,
+      "step": 3604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9768603883418204e-05,
+      "loss": 1.8413,
+      "step": 3605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.976847563244931e-05,
+      "loss": 2.1207,
+      "step": 3606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9768347346365057e-05,
+      "loss": 1.8672,
+      "step": 3607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9768219025165908e-05,
+      "loss": 2.1457,
+      "step": 3608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9768090668852317e-05,
+      "loss": 1.9291,
+      "step": 3609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.976796227742475e-05,
+      "loss": 2.1134,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.976783385088367e-05,
+      "loss": 1.7605,
+      "step": 3611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9767705389229536e-05,
+      "loss": 1.9673,
+      "step": 3612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9767576892462815e-05,
+      "loss": 2.1203,
+      "step": 3613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.976744836058396e-05,
+      "loss": 2.1982,
+      "step": 3614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.976731979359344e-05,
+      "loss": 2.0324,
+      "step": 3615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9767191191491718e-05,
+      "loss": 1.717,
+      "step": 3616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9767062554279254e-05,
+      "loss": 1.5716,
+      "step": 3617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9766933881956504e-05,
+      "loss": 1.8398,
+      "step": 3618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.976680517452394e-05,
+      "loss": 1.8546,
+      "step": 3619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9766676431982022e-05,
+      "loss": 1.8324,
+      "step": 3620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.976654765433121e-05,
+      "loss": 2.1695,
+      "step": 3621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.976641884157197e-05,
+      "loss": 2.1189,
+      "step": 3622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9766289993704763e-05,
+      "loss": 1.86,
+      "step": 3623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9766161110730056e-05,
+      "loss": 1.752,
+      "step": 3624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9766032192648307e-05,
+      "loss": 1.7896,
+      "step": 3625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.976590323945998e-05,
+      "loss": 1.6348,
+      "step": 3626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9765774251165545e-05,
+      "loss": 1.9004,
+      "step": 3627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.976564522776546e-05,
+      "loss": 2.0273,
+      "step": 3628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9765516169260184e-05,
+      "loss": 1.99,
+      "step": 3629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9765387075650193e-05,
+      "loss": 1.8629,
+      "step": 3630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9765257946935944e-05,
+      "loss": 1.9115,
+      "step": 3631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9765128783117903e-05,
+      "loss": 1.9083,
+      "step": 3632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9764999584196532e-05,
+      "loss": 1.8385,
+      "step": 3633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.97648703501723e-05,
+      "loss": 1.6344,
+      "step": 3634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.976474108104566e-05,
+      "loss": 1.6672,
+      "step": 3635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.976461177681709e-05,
+      "loss": 1.7991,
+      "step": 3636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9764482437487053e-05,
+      "loss": 1.8307,
+      "step": 3637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9764353063056007e-05,
+      "loss": 1.6609,
+      "step": 3638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9764223653524424e-05,
+      "loss": 2.1105,
+      "step": 3639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9764094208892765e-05,
+      "loss": 2.0776,
+      "step": 3640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9763964729161493e-05,
+      "loss": 2.3151,
+      "step": 3641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.976383521433108e-05,
+      "loss": 1.9876,
+      "step": 3642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9763705664401987e-05,
+      "loss": 1.9555,
+      "step": 3643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9763576079374684e-05,
+      "loss": 2.0265,
+      "step": 3644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.976344645924963e-05,
+      "loss": 1.8831,
+      "step": 3645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.97633168040273e-05,
+      "loss": 1.9152,
+      "step": 3646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9763187113708154e-05,
+      "loss": 1.6696,
+      "step": 3647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9763057388292656e-05,
+      "loss": 1.8486,
+      "step": 3648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.9762927627781278e-05,
+      "loss": 2.2817,
+      "step": 3649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9762797832174483e-05,
+      "loss": 1.906,
+      "step": 3650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.976266800147274e-05,
+      "loss": 1.5876,
+      "step": 3651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.976253813567651e-05,
+      "loss": 1.6855,
+      "step": 3652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.976240823478627e-05,
+      "loss": 1.8671,
+      "step": 3653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.976227829880248e-05,
+      "loss": 1.8348,
+      "step": 3654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9762148327725605e-05,
+      "loss": 1.9283,
+      "step": 3655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9762018321556117e-05,
+      "loss": 1.9753,
+      "step": 3656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.976188828029448e-05,
+      "loss": 1.4119,
+      "step": 3657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.976175820394117e-05,
+      "loss": 2.0283,
+      "step": 3658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.976162809249664e-05,
+      "loss": 1.6859,
+      "step": 3659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.976149794596137e-05,
+      "loss": 1.8793,
+      "step": 3660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.976136776433582e-05,
+      "loss": 1.9796,
+      "step": 3661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9761237547620466e-05,
+      "loss": 1.6363,
+      "step": 3662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9761107295815766e-05,
+      "loss": 2.052,
+      "step": 3663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9760977008922198e-05,
+      "loss": 1.8487,
+      "step": 3664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9760846686940222e-05,
+      "loss": 1.6494,
+      "step": 3665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9760716329870316e-05,
+      "loss": 2.1564,
+      "step": 3666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9760585937712938e-05,
+      "loss": 1.7639,
+      "step": 3667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9760455510468565e-05,
+      "loss": 1.9497,
+      "step": 3668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.875,
+      "learning_rate": 1.9760325048137665e-05,
+      "loss": 3.2551,
+      "step": 3669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9760194550720698e-05,
+      "loss": 1.6434,
+      "step": 3670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9760064018218146e-05,
+      "loss": 2.0268,
+      "step": 3671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.975993345063047e-05,
+      "loss": 2.3091,
+      "step": 3672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.975980284795814e-05,
+      "loss": 2.025,
+      "step": 3673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9759672210201627e-05,
+      "loss": 1.5566,
+      "step": 3674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9759541537361403e-05,
+      "loss": 2.4874,
+      "step": 3675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.975941082943793e-05,
+      "loss": 1.8779,
+      "step": 3676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.975928008643169e-05,
+      "loss": 1.7874,
+      "step": 3677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9759149308343143e-05,
+      "loss": 2.1805,
+      "step": 3678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.975901849517276e-05,
+      "loss": 1.7743,
+      "step": 3679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9758887646921016e-05,
+      "loss": 1.6229,
+      "step": 3680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9758756763588378e-05,
+      "loss": 1.9113,
+      "step": 3681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9758625845175315e-05,
+      "loss": 2.1866,
+      "step": 3682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9758494891682308e-05,
+      "loss": 1.9602,
+      "step": 3683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.375,
+      "learning_rate": 1.975836390310981e-05,
+      "loss": 1.7406,
+      "step": 3684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9758232879458308e-05,
+      "loss": 1.7256,
+      "step": 3685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9758101820728264e-05,
+      "loss": 2.3534,
+      "step": 3686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9757970726920153e-05,
+      "loss": 1.7669,
+      "step": 3687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9757839598034442e-05,
+      "loss": 1.8592,
+      "step": 3688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9757708434071606e-05,
+      "loss": 2.033,
+      "step": 3689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.975757723503212e-05,
+      "loss": 2.0312,
+      "step": 3690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9757446000916446e-05,
+      "loss": 2.1104,
+      "step": 3691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9757314731725064e-05,
+      "loss": 2.3818,
+      "step": 3692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9757183427458444e-05,
+      "loss": 2.026,
+      "step": 3693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9757052088117054e-05,
+      "loss": 1.9931,
+      "step": 3694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9756920713701374e-05,
+      "loss": 1.9753,
+      "step": 3695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9756789304211866e-05,
+      "loss": 2.1682,
+      "step": 3696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.975665785964901e-05,
+      "loss": 1.9856,
+      "step": 3697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9756526380013276e-05,
+      "loss": 1.8622,
+      "step": 3698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9756394865305138e-05,
+      "loss": 2.3299,
+      "step": 3699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.975626331552507e-05,
+      "loss": 1.8854,
+      "step": 3700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.975613173067354e-05,
+      "loss": 2.2535,
+      "step": 3701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9756000110751023e-05,
+      "loss": 1.9381,
+      "step": 3702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9755868455757998e-05,
+      "loss": 1.9073,
+      "step": 3703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9755736765694924e-05,
+      "loss": 2.3588,
+      "step": 3704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.975560504056229e-05,
+      "loss": 1.8571,
+      "step": 3705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.975547328036056e-05,
+      "loss": 1.7501,
+      "step": 3706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9755341485090215e-05,
+      "loss": 1.7751,
+      "step": 3707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.975520965475172e-05,
+      "loss": 1.9993,
+      "step": 3708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9755077789345555e-05,
+      "loss": 1.9914,
+      "step": 3709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.975494588887219e-05,
+      "loss": 1.9345,
+      "step": 3710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9754813953332107e-05,
+      "loss": 1.7727,
+      "step": 3711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.975468198272577e-05,
+      "loss": 1.9932,
+      "step": 3712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.975454997705366e-05,
+      "loss": 1.8633,
+      "step": 3713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.975441793631625e-05,
+      "loss": 1.7561,
+      "step": 3714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9754285860514012e-05,
+      "loss": 2.0806,
+      "step": 3715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9754153749647424e-05,
+      "loss": 2.0911,
+      "step": 3716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.975402160371696e-05,
+      "loss": 1.738,
+      "step": 3717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.97538894227231e-05,
+      "loss": 2.011,
+      "step": 3718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9753757206666308e-05,
+      "loss": 2.1032,
+      "step": 3719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.975362495554707e-05,
+      "loss": 2.0992,
+      "step": 3720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9753492669365853e-05,
+      "loss": 1.676,
+      "step": 3721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.975336034812314e-05,
+      "loss": 2.0113,
+      "step": 3722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9753227991819402e-05,
+      "loss": 1.9975,
+      "step": 3723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.9753095600455112e-05,
+      "loss": 2.7578,
+      "step": 3724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9752963174030753e-05,
+      "loss": 2.2758,
+      "step": 3725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.97528307125468e-05,
+      "loss": 2.0272,
+      "step": 3726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9752698216003724e-05,
+      "loss": 2.1001,
+      "step": 3727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9752565684402007e-05,
+      "loss": 1.9672,
+      "step": 3728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.975243311774212e-05,
+      "loss": 1.9332,
+      "step": 3729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9752300516024544e-05,
+      "loss": 1.8813,
+      "step": 3730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9752167879249754e-05,
+      "loss": 1.8324,
+      "step": 3731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.975203520741823e-05,
+      "loss": 1.7322,
+      "step": 3732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.975190250053044e-05,
+      "loss": 1.7774,
+      "step": 3733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9751769758586866e-05,
+      "loss": 1.6556,
+      "step": 3734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.975163698158799e-05,
+      "loss": 1.8791,
+      "step": 3735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9751504169534285e-05,
+      "loss": 1.8378,
+      "step": 3736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9751371322426228e-05,
+      "loss": 1.5655,
+      "step": 3737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.97512384402643e-05,
+      "loss": 2.0493,
+      "step": 3738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.975110552304897e-05,
+      "loss": 1.9642,
+      "step": 3739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9750972570780725e-05,
+      "loss": 1.9696,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.975083958346004e-05,
+      "loss": 1.7366,
+      "step": 3741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.975070656108739e-05,
+      "loss": 1.8126,
+      "step": 3742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.975057350366326e-05,
+      "loss": 1.7992,
+      "step": 3743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9750440411188122e-05,
+      "loss": 2.1201,
+      "step": 3744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9750307283662456e-05,
+      "loss": 1.7575,
+      "step": 3745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9750174121086743e-05,
+      "loss": 1.8943,
+      "step": 3746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9750040923461457e-05,
+      "loss": 1.7564,
+      "step": 3747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.974990769078708e-05,
+      "loss": 1.9842,
+      "step": 3748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9749774423064094e-05,
+      "loss": 1.8707,
+      "step": 3749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9749641120292973e-05,
+      "loss": 1.769,
+      "step": 3750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9749507782474197e-05,
+      "loss": 1.91,
+      "step": 3751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9749374409608247e-05,
+      "loss": 2.0134,
+      "step": 3752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9749241001695603e-05,
+      "loss": 2.4378,
+      "step": 3753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9749107558736738e-05,
+      "loss": 1.5484,
+      "step": 3754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9748974080732142e-05,
+      "loss": 1.9187,
+      "step": 3755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.974884056768229e-05,
+      "loss": 1.9594,
+      "step": 3756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.974870701958766e-05,
+      "loss": 1.8263,
+      "step": 3757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9748573436448734e-05,
+      "loss": 1.9759,
+      "step": 3758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9748439818265993e-05,
+      "loss": 1.9148,
+      "step": 3759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9748306165039913e-05,
+      "loss": 1.3463,
+      "step": 3760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9748172476770982e-05,
+      "loss": 1.9073,
+      "step": 3761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9748038753459675e-05,
+      "loss": 1.916,
+      "step": 3762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9747904995106472e-05,
+      "loss": 1.8067,
+      "step": 3763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.974777120171186e-05,
+      "loss": 1.7914,
+      "step": 3764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.974763737327631e-05,
+      "loss": 1.7534,
+      "step": 3765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9747503509800313e-05,
+      "loss": 1.9838,
+      "step": 3766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9747369611284344e-05,
+      "loss": 1.9835,
+      "step": 3767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9747235677728888e-05,
+      "loss": 2.0456,
+      "step": 3768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9747101709134427e-05,
+      "loss": 1.9353,
+      "step": 3769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9746967705501435e-05,
+      "loss": 2.1162,
+      "step": 3770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9746833666830403e-05,
+      "loss": 1.7784,
+      "step": 3771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9746699593121807e-05,
+      "loss": 1.7404,
+      "step": 3772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9746565484376132e-05,
+      "loss": 2.0364,
+      "step": 3773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.974643134059386e-05,
+      "loss": 1.8911,
+      "step": 3774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9746297161775473e-05,
+      "loss": 1.73,
+      "step": 3775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.974616294792145e-05,
+      "loss": 1.8396,
+      "step": 3776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9746028699032277e-05,
+      "loss": 1.9677,
+      "step": 3777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9745894415108435e-05,
+      "loss": 1.9547,
+      "step": 3778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9745760096150407e-05,
+      "loss": 1.7424,
+      "step": 3779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9745625742158682e-05,
+      "loss": 2.0139,
+      "step": 3780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.974549135313373e-05,
+      "loss": 1.8004,
+      "step": 3781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9745356929076042e-05,
+      "loss": 1.9643,
+      "step": 3782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9745222469986106e-05,
+      "loss": 1.9628,
+      "step": 3783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9745087975864395e-05,
+      "loss": 1.6412,
+      "step": 3784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.97449534467114e-05,
+      "loss": 1.9367,
+      "step": 3785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.97448188825276e-05,
+      "loss": 1.8662,
+      "step": 3786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9744684283313483e-05,
+      "loss": 1.7809,
+      "step": 3787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9744549649069527e-05,
+      "loss": 1.5327,
+      "step": 3788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9744414979796222e-05,
+      "loss": 1.9382,
+      "step": 3789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.974428027549405e-05,
+      "loss": 1.8743,
+      "step": 3790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9744145536163496e-05,
+      "loss": 2.0041,
+      "step": 3791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.974401076180504e-05,
+      "loss": 2.2319,
+      "step": 3792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9743875952419168e-05,
+      "loss": 1.9895,
+      "step": 3793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9743741108006375e-05,
+      "loss": 1.7028,
+      "step": 3794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.974360622856713e-05,
+      "loss": 2.0345,
+      "step": 3795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9743471314101927e-05,
+      "loss": 1.8599,
+      "step": 3796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9743336364611244e-05,
+      "loss": 1.9929,
+      "step": 3797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9743201380095576e-05,
+      "loss": 1.8519,
+      "step": 3798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9743066360555404e-05,
+      "loss": 2.0022,
+      "step": 3799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.974293130599121e-05,
+      "loss": 1.9399,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9742796216403483e-05,
+      "loss": 1.702,
+      "step": 3801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.974266109179271e-05,
+      "loss": 1.7769,
+      "step": 3802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9742525932159373e-05,
+      "loss": 1.7272,
+      "step": 3803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.974239073750396e-05,
+      "loss": 1.8103,
+      "step": 3804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9742255507826954e-05,
+      "loss": 2.0556,
+      "step": 3805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9742120243128844e-05,
+      "loss": 1.8941,
+      "step": 3806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9741984943410116e-05,
+      "loss": 1.7695,
+      "step": 3807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9741849608671254e-05,
+      "loss": 2.0069,
+      "step": 3808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.974171423891275e-05,
+      "loss": 1.7559,
+      "step": 3809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9741578834135086e-05,
+      "loss": 1.6865,
+      "step": 3810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9741443394338748e-05,
+      "loss": 1.556,
+      "step": 3811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.974130791952423e-05,
+      "loss": 1.9,
+      "step": 3812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9741172409692008e-05,
+      "loss": 1.8911,
+      "step": 3813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.974103686484258e-05,
+      "loss": 1.9101,
+      "step": 3814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9740901284976424e-05,
+      "loss": 2.0548,
+      "step": 3815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9740765670094033e-05,
+      "loss": 1.5229,
+      "step": 3816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9740630020195892e-05,
+      "loss": 1.8427,
+      "step": 3817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.974049433528249e-05,
+      "loss": 2.0393,
+      "step": 3818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9740358615354315e-05,
+      "loss": 1.9897,
+      "step": 3819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9740222860411854e-05,
+      "loss": 1.941,
+      "step": 3820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9740087070455596e-05,
+      "loss": 1.9921,
+      "step": 3821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.973995124548603e-05,
+      "loss": 2.0281,
+      "step": 3822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.973981538550364e-05,
+      "loss": 1.683,
+      "step": 3823
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.973967949050892e-05,
+      "loss": 1.6016,
+      "step": 3824
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9739543560502352e-05,
+      "loss": 1.9151,
+      "step": 3825
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9739407595484428e-05,
+      "loss": 1.9573,
+      "step": 3826
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.973927159545564e-05,
+      "loss": 1.9378,
+      "step": 3827
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.973913556041647e-05,
+      "loss": 1.8148,
+      "step": 3828
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9738999490367417e-05,
+      "loss": 1.899,
+      "step": 3829
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.973886338530896e-05,
+      "loss": 1.7857,
+      "step": 3830
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.973872724524159e-05,
+      "loss": 2.2474,
+      "step": 3831
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9738591070165804e-05,
+      "loss": 1.4754,
+      "step": 3832
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9738454860082084e-05,
+      "loss": 1.9876,
+      "step": 3833
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9738318614990924e-05,
+      "loss": 1.8239,
+      "step": 3834
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9738182334892806e-05,
+      "loss": 2.2787,
+      "step": 3835
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9738046019788227e-05,
+      "loss": 1.6871,
+      "step": 3836
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.973790966967768e-05,
+      "loss": 1.9845,
+      "step": 3837
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.973777328456165e-05,
+      "loss": 1.9732,
+      "step": 3838
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9737636864440627e-05,
+      "loss": 1.8404,
+      "step": 3839
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9737500409315098e-05,
+      "loss": 1.8309,
+      "step": 3840
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9737363919185562e-05,
+      "loss": 1.8202,
+      "step": 3841
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9737227394052506e-05,
+      "loss": 1.8377,
+      "step": 3842
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9737090833916416e-05,
+      "loss": 2.0398,
+      "step": 3843
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9736954238777793e-05,
+      "loss": 1.8784,
+      "step": 3844
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9736817608637117e-05,
+      "loss": 2.1906,
+      "step": 3845
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9736680943494888e-05,
+      "loss": 1.9601,
+      "step": 3846
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9736544243351594e-05,
+      "loss": 1.6961,
+      "step": 3847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.9736407508207726e-05,
+      "loss": 2.7093,
+      "step": 3848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9736270738063772e-05,
+      "loss": 2.076,
+      "step": 3849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.973613393292023e-05,
+      "loss": 2.5077,
+      "step": 3850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.973599709277759e-05,
+      "loss": 1.6873,
+      "step": 3851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.973586021763634e-05,
+      "loss": 1.892,
+      "step": 3852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9735723307496976e-05,
+      "loss": 1.7838,
+      "step": 3853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.973558636235999e-05,
+      "loss": 1.9821,
+      "step": 3854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9735449382225872e-05,
+      "loss": 2.1473,
+      "step": 3855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9735312367095118e-05,
+      "loss": 1.791,
+      "step": 3856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9735175316968215e-05,
+      "loss": 1.8289,
+      "step": 3857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.9735038231845664e-05,
+      "loss": 2.0585,
+      "step": 3858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.973490111172795e-05,
+      "loss": 1.6414,
+      "step": 3859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9734763956615566e-05,
+      "loss": 1.8108,
+      "step": 3860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.973462676650901e-05,
+      "loss": 1.8233,
+      "step": 3861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9734489541408778e-05,
+      "loss": 2.209,
+      "step": 3862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9734352281315353e-05,
+      "loss": 2.0016,
+      "step": 3863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9734214986229236e-05,
+      "loss": 1.9414,
+      "step": 3864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9734077656150913e-05,
+      "loss": 1.9383,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9733940291080887e-05,
+      "loss": 1.9984,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.973380289101965e-05,
+      "loss": 1.6904,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.973366545596769e-05,
+      "loss": 2.2025,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9733527985925504e-05,
+      "loss": 1.9812,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.973339048089359e-05,
+      "loss": 2.4124,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9733252940872436e-05,
+      "loss": 1.983,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.973311536586254e-05,
+      "loss": 1.9862,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.97329777558644e-05,
+      "loss": 2.0537,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9732840110878505e-05,
+      "loss": 1.7491,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.973270243090535e-05,
+      "loss": 1.9829,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9732564715945427e-05,
+      "loss": 1.8897,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9732426965999242e-05,
+      "loss": 1.7048,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9732289181067283e-05,
+      "loss": 1.7886,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9732151361150044e-05,
+      "loss": 2.0364,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9732013506248024e-05,
+      "loss": 1.9684,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9731875616361717e-05,
+      "loss": 1.6814,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.973173769149162e-05,
+      "loss": 1.8604,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9731599731638223e-05,
+      "loss": 2.1206,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.973146173680203e-05,
+      "loss": 1.9794,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9731323706983532e-05,
+      "loss": 1.9915,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9731185642183227e-05,
+      "loss": 1.8961,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.973104754240161e-05,
+      "loss": 1.7343,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.973090940763918e-05,
+      "loss": 2.1916,
+      "step": 3888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9730771237896432e-05,
+      "loss": 1.9383,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9730633033173863e-05,
+      "loss": 1.6114,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9730494793471968e-05,
+      "loss": 1.7538,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9730356518791247e-05,
+      "loss": 1.8401,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9730218209132196e-05,
+      "loss": 2.0222,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.973007986449531e-05,
+      "loss": 1.8677,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9729941484881084e-05,
+      "loss": 2.1769,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9729803070290027e-05,
+      "loss": 1.8576,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9729664620722623e-05,
+      "loss": 2.1042,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9729526136179383e-05,
+      "loss": 1.5476,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.972938761666079e-05,
+      "loss": 1.9961,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.972924906216735e-05,
+      "loss": 2.1429,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9729110472699565e-05,
+      "loss": 1.9502,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9728971848257927e-05,
+      "loss": 1.7872,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9728833188842936e-05,
+      "loss": 1.592,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9728694494455093e-05,
+      "loss": 2.1154,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9728555765094886e-05,
+      "loss": 2.123,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.972841700076283e-05,
+      "loss": 2.182,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9728278201459413e-05,
+      "loss": 1.8364,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9728139367185135e-05,
+      "loss": 2.3551,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9728000497940497e-05,
+      "loss": 2.0173,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9727861593725997e-05,
+      "loss": 2.1638,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.972772265454214e-05,
+      "loss": 2.0427,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9727583680389414e-05,
+      "loss": 1.911,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.972744467126833e-05,
+      "loss": 2.1897,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.972730562717938e-05,
+      "loss": 2.0631,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.972716654812307e-05,
+      "loss": 1.9592,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9727027434099896e-05,
+      "loss": 1.7259,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9726888285110358e-05,
+      "loss": 2.1198,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9726749101154954e-05,
+      "loss": 1.7062,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.972660988223419e-05,
+      "loss": 2.0031,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9726470628348566e-05,
+      "loss": 2.0231,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9726331339498577e-05,
+      "loss": 1.8418,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9726192015684725e-05,
+      "loss": 1.9693,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9726052656907518e-05,
+      "loss": 1.6375,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9725913263167447e-05,
+      "loss": 1.6109,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.972577383446502e-05,
+      "loss": 2.0017,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9725634370800737e-05,
+      "loss": 1.7605,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.97254948721751e-05,
+      "loss": 1.8242,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9725355338588606e-05,
+      "loss": 1.6054,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9725215770041762e-05,
+      "loss": 2.2446,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9725076166535063e-05,
+      "loss": 1.9847,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.972493652806902e-05,
+      "loss": 1.7991,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9724796854644128e-05,
+      "loss": 1.8992,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9724657146260892e-05,
+      "loss": 2.1886,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.972451740291981e-05,
+      "loss": 1.4819,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.972437762462139e-05,
+      "loss": 1.677,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9724237811366133e-05,
+      "loss": 2.1373,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.972409796315454e-05,
+      "loss": 1.7935,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9723958079987118e-05,
+      "loss": 1.8993,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9723818161864362e-05,
+      "loss": 1.7996,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.972367820878678e-05,
+      "loss": 2.353,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9723538220754874e-05,
+      "loss": 2.1559,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.972339819776915e-05,
+      "loss": 2.0662,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9723258139830106e-05,
+      "loss": 2.1684,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.972311804693825e-05,
+      "loss": 1.9304,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9722977919094087e-05,
+      "loss": 2.0034,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9722837756298112e-05,
+      "loss": 1.6345,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9722697558550837e-05,
+      "loss": 1.9127,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9722557325852767e-05,
+      "loss": 2.443,
+      "step": 3948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.97224170582044e-05,
+      "loss": 1.8685,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.972227675560624e-05,
+      "loss": 1.8386,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9722136418058797e-05,
+      "loss": 2.305,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9721996045562574e-05,
+      "loss": 1.9253,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9721855638118073e-05,
+      "loss": 1.797,
+      "step": 3953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9721715195725803e-05,
+      "loss": 1.6967,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9721574718386264e-05,
+      "loss": 2.0923,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.972143420609996e-05,
+      "loss": 1.8949,
+      "step": 3956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9721293658867398e-05,
+      "loss": 1.6753,
+      "step": 3957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.972115307668909e-05,
+      "loss": 2.027,
+      "step": 3958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9721012459565534e-05,
+      "loss": 2.1269,
+      "step": 3959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9720871807497237e-05,
+      "loss": 1.8919,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.9720731120484705e-05,
+      "loss": 1.2957,
+      "step": 3961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9720590398528444e-05,
+      "loss": 1.9258,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9720449641628958e-05,
+      "loss": 2.3684,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9720308849786753e-05,
+      "loss": 1.9084,
+      "step": 3964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.972016802300234e-05,
+      "loss": 1.9898,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.972002716127622e-05,
+      "loss": 1.7427,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.97198862646089e-05,
+      "loss": 1.8758,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.971974533300089e-05,
+      "loss": 1.8037,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.971960436645269e-05,
+      "loss": 1.8189,
+      "step": 3969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9719463364964815e-05,
+      "loss": 1.6901,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9719322328537768e-05,
+      "loss": 1.8484,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9719181257172055e-05,
+      "loss": 2.0431,
+      "step": 3972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.971904015086818e-05,
+      "loss": 1.6493,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9718899009626657e-05,
+      "loss": 1.962,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9718757833447993e-05,
+      "loss": 2.0139,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.971861662233269e-05,
+      "loss": 2.0215,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9718475376281257e-05,
+      "loss": 2.0433,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.971833409529421e-05,
+      "loss": 2.0156,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9718192779372042e-05,
+      "loss": 1.7898,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9718051428515275e-05,
+      "loss": 1.7702,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.971791004272441e-05,
+      "loss": 2.0532,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9717768621999954e-05,
+      "loss": 1.837,
+      "step": 3982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.971762716634242e-05,
+      "loss": 2.0175,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9717485675752316e-05,
+      "loss": 1.7287,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.971734415023015e-05,
+      "loss": 1.7466,
+      "step": 3985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9717202589776424e-05,
+      "loss": 1.8064,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9717060994391657e-05,
+      "loss": 1.8975,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9716919364076356e-05,
+      "loss": 1.7567,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9716777698831026e-05,
+      "loss": 1.8185,
+      "step": 3989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9716635998656174e-05,
+      "loss": 2.0586,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9716494263552316e-05,
+      "loss": 1.9827,
+      "step": 3991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.971635249351996e-05,
+      "loss": 1.8458,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9716210688559615e-05,
+      "loss": 1.9938,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9716068848671788e-05,
+      "loss": 1.9655,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9715926973856995e-05,
+      "loss": 1.7036,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9715785064115743e-05,
+      "loss": 1.8391,
+      "step": 3996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9715643119448534e-05,
+      "loss": 2.1357,
+      "step": 3997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9715501139855894e-05,
+      "loss": 1.9314,
+      "step": 3998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.971535912533832e-05,
+      "loss": 1.6609,
+      "step": 3999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.9715217075896328e-05,
+      "loss": 2.2449,
+      "step": 4000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9715074991530428e-05,
+      "loss": 1.9323,
+      "step": 4001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9714932872241136e-05,
+      "loss": 1.7326,
+      "step": 4002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9714790718028954e-05,
+      "loss": 2.0993,
+      "step": 4003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9714648528894396e-05,
+      "loss": 2.1934,
+      "step": 4004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9714506304837975e-05,
+      "loss": 2.0763,
+      "step": 4005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.97143640458602e-05,
+      "loss": 1.7107,
+      "step": 4006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9714221751961586e-05,
+      "loss": 1.5769,
+      "step": 4007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.971407942314264e-05,
+      "loss": 2.0984,
+      "step": 4008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9713937059403876e-05,
+      "loss": 1.9989,
+      "step": 4009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9713794660745805e-05,
+      "loss": 1.8144,
+      "step": 4010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9713652227168944e-05,
+      "loss": 2.0913,
+      "step": 4011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9713509758673795e-05,
+      "loss": 1.8149,
+      "step": 4012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9713367255260876e-05,
+      "loss": 1.8964,
+      "step": 4013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9713224716930704e-05,
+      "loss": 1.8689,
+      "step": 4014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.971308214368378e-05,
+      "loss": 1.7967,
+      "step": 4015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.971293953552063e-05,
+      "loss": 1.9272,
+      "step": 4016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.971279689244175e-05,
+      "loss": 1.871,
+      "step": 4017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.971265421444767e-05,
+      "loss": 1.9676,
+      "step": 4018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9712511501538892e-05,
+      "loss": 2.3031,
+      "step": 4019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9712368753715934e-05,
+      "loss": 1.7439,
+      "step": 4020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9712225970979305e-05,
+      "loss": 1.7588,
+      "step": 4021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.971208315332952e-05,
+      "loss": 1.9582,
+      "step": 4022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9711940300767094e-05,
+      "loss": 1.9078,
+      "step": 4023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.971179741329254e-05,
+      "loss": 1.9155,
+      "step": 4024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9711654490906368e-05,
+      "loss": 2.3186,
+      "step": 4025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.97115115336091e-05,
+      "loss": 1.7296,
+      "step": 4026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9711368541401242e-05,
+      "loss": 1.9767,
+      "step": 4027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9711225514283308e-05,
+      "loss": 2.0847,
+      "step": 4028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.971108245225582e-05,
+      "loss": 2.1368,
+      "step": 4029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.9710939355319284e-05,
+      "loss": 1.8574,
+      "step": 4030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9710796223474217e-05,
+      "loss": 1.8866,
+      "step": 4031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.971065305672114e-05,
+      "loss": 1.85,
+      "step": 4032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9710509855060555e-05,
+      "loss": 2.0376,
+      "step": 4033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9710366618492986e-05,
+      "loss": 1.8976,
+      "step": 4034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9710223347018945e-05,
+      "loss": 1.809,
+      "step": 4035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.971008004063895e-05,
+      "loss": 2.1099,
+      "step": 4036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.970993669935351e-05,
+      "loss": 1.8305,
+      "step": 4037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9709793323163144e-05,
+      "loss": 1.9577,
+      "step": 4038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.970964991206837e-05,
+      "loss": 2.292,
+      "step": 4039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9709506466069703e-05,
+      "loss": 2.1407,
+      "step": 4040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.970936298516765e-05,
+      "loss": 1.936,
+      "step": 4041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9709219469362738e-05,
+      "loss": 2.3755,
+      "step": 4042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9709075918655474e-05,
+      "loss": 1.9226,
+      "step": 4043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9708932333046383e-05,
+      "loss": 1.9481,
+      "step": 4044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9708788712535974e-05,
+      "loss": 2.4435,
+      "step": 4045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9708645057124765e-05,
+      "loss": 1.8261,
+      "step": 4046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.970850136681327e-05,
+      "loss": 2.0452,
+      "step": 4047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9708357641602018e-05,
+      "loss": 1.7414,
+      "step": 4048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9708213881491506e-05,
+      "loss": 2.1221,
+      "step": 4049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.970807008648227e-05,
+      "loss": 1.9671,
+      "step": 4050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.970792625657481e-05,
+      "loss": 1.769,
+      "step": 4051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9707782391769657e-05,
+      "loss": 1.984,
+      "step": 4052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.970763849206732e-05,
+      "loss": 1.7607,
+      "step": 4053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9707494557468315e-05,
+      "loss": 2.0065,
+      "step": 4054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.970735058797317e-05,
+      "loss": 1.8494,
+      "step": 4055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9707206583582387e-05,
+      "loss": 1.7776,
+      "step": 4056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9707062544296497e-05,
+      "loss": 1.9023,
+      "step": 4057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9706918470116013e-05,
+      "loss": 2.1299,
+      "step": 4058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9706774361041453e-05,
+      "loss": 1.9624,
+      "step": 4059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9706630217073332e-05,
+      "loss": 1.957,
+      "step": 4060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9706486038212174e-05,
+      "loss": 1.858,
+      "step": 4061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9706341824458493e-05,
+      "loss": 1.6575,
+      "step": 4062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9706197575812807e-05,
+      "loss": 1.7289,
+      "step": 4063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9706053292275636e-05,
+      "loss": 2.4555,
+      "step": 4064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9705908973847504e-05,
+      "loss": 1.755,
+      "step": 4065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.970576462052892e-05,
+      "loss": 2.0719,
+      "step": 4066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9705620232320407e-05,
+      "loss": 2.2156,
+      "step": 4067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9705475809222485e-05,
+      "loss": 1.9527,
+      "step": 4068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9705331351235673e-05,
+      "loss": 1.5674,
+      "step": 4069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9705186858360492e-05,
+      "loss": 1.5708,
+      "step": 4070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.970504233059746e-05,
+      "loss": 1.8844,
+      "step": 4071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9704897767947094e-05,
+      "loss": 1.6142,
+      "step": 4072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.9704753170409913e-05,
+      "loss": 2.3915,
+      "step": 4073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9704608537986444e-05,
+      "loss": 1.9491,
+      "step": 4074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.97044638706772e-05,
+      "loss": 1.9498,
+      "step": 4075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9704319168482703e-05,
+      "loss": 1.9151,
+      "step": 4076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9704174431403474e-05,
+      "loss": 1.9917,
+      "step": 4077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9704029659440034e-05,
+      "loss": 1.931,
+      "step": 4078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9703884852592904e-05,
+      "loss": 2.0247,
+      "step": 4079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9703740010862603e-05,
+      "loss": 1.8078,
+      "step": 4080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9703595134249645e-05,
+      "loss": 1.9192,
+      "step": 4081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9703450222754563e-05,
+      "loss": 2.0901,
+      "step": 4082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9703305276377874e-05,
+      "loss": 1.8152,
+      "step": 4083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9703160295120094e-05,
+      "loss": 1.941,
+      "step": 4084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9703015278981748e-05,
+      "loss": 1.8337,
+      "step": 4085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9702870227963355e-05,
+      "loss": 1.8875,
+      "step": 4086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.970272514206544e-05,
+      "loss": 1.9658,
+      "step": 4087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9702580021288522e-05,
+      "loss": 1.7638,
+      "step": 4088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9702434865633126e-05,
+      "loss": 1.8054,
+      "step": 4089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.970228967509977e-05,
+      "loss": 1.8547,
+      "step": 4090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9702144449688974e-05,
+      "loss": 1.8894,
+      "step": 4091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9701999189401266e-05,
+      "loss": 1.6017,
+      "step": 4092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9701853894237165e-05,
+      "loss": 1.8194,
+      "step": 4093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9701708564197194e-05,
+      "loss": 1.5906,
+      "step": 4094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9701563199281872e-05,
+      "loss": 2.147,
+      "step": 4095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.970141779949173e-05,
+      "loss": 2.0144,
+      "step": 4096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.970127236482728e-05,
+      "loss": 1.7974,
+      "step": 4097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9701126895289053e-05,
+      "loss": 1.9583,
+      "step": 4098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9700981390877568e-05,
+      "loss": 1.7638,
+      "step": 4099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.970083585159335e-05,
+      "loss": 1.8047,
+      "step": 4100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9700690277436923e-05,
+      "loss": 1.9106,
+      "step": 4101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9700544668408806e-05,
+      "loss": 1.9648,
+      "step": 4102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9700399024509528e-05,
+      "loss": 2.0667,
+      "step": 4103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9700253345739604e-05,
+      "loss": 2.0558,
+      "step": 4104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9700107632099567e-05,
+      "loss": 2.1054,
+      "step": 4105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9699961883589938e-05,
+      "loss": 1.9263,
+      "step": 4106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.969981610021124e-05,
+      "loss": 1.9921,
+      "step": 4107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9699670281963995e-05,
+      "loss": 1.8966,
+      "step": 4108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9699524428848728e-05,
+      "loss": 1.8895,
+      "step": 4109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9699378540865966e-05,
+      "loss": 1.8301,
+      "step": 4110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.969923261801623e-05,
+      "loss": 1.8473,
+      "step": 4111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9699086660300052e-05,
+      "loss": 1.8443,
+      "step": 4112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.969894066771795e-05,
+      "loss": 1.9572,
+      "step": 4113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9698794640270445e-05,
+      "loss": 1.9218,
+      "step": 4114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9698648577958074e-05,
+      "loss": 2.1676,
+      "step": 4115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9698502480781352e-05,
+      "loss": 1.9704,
+      "step": 4116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9698356348740802e-05,
+      "loss": 1.9194,
+      "step": 4117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9698210181836962e-05,
+      "loss": 1.9886,
+      "step": 4118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9698063980070347e-05,
+      "loss": 1.7081,
+      "step": 4119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9697917743441485e-05,
+      "loss": 2.2438,
+      "step": 4120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9697771471950906e-05,
+      "loss": 1.7034,
+      "step": 4121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9697625165599127e-05,
+      "loss": 2.0541,
+      "step": 4122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.969747882438668e-05,
+      "loss": 1.9314,
+      "step": 4123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.969733244831409e-05,
+      "loss": 1.7516,
+      "step": 4124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9697186037381885e-05,
+      "loss": 1.7266,
+      "step": 4125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9697039591590587e-05,
+      "loss": 2.1449,
+      "step": 4126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.969689311094073e-05,
+      "loss": 2.3255,
+      "step": 4127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9696746595432828e-05,
+      "loss": 1.805,
+      "step": 4128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.969660004506742e-05,
+      "loss": 1.7182,
+      "step": 4129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9696453459845027e-05,
+      "loss": 2.3001,
+      "step": 4130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9696306839766177e-05,
+      "loss": 1.7044,
+      "step": 4131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.96961601848314e-05,
+      "loss": 2.0271,
+      "step": 4132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9696013495041214e-05,
+      "loss": 1.8874,
+      "step": 4133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9695866770396157e-05,
+      "loss": 2.0214,
+      "step": 4134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.969572001089675e-05,
+      "loss": 1.913,
+      "step": 4135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9695573216543523e-05,
+      "loss": 1.8642,
+      "step": 4136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9695426387337005e-05,
+      "loss": 2.0265,
+      "step": 4137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.969527952327772e-05,
+      "loss": 2.1013,
+      "step": 4138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.96951326243662e-05,
+      "loss": 1.9615,
+      "step": 4139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.969498569060297e-05,
+      "loss": 2.0124,
+      "step": 4140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9694838721988557e-05,
+      "loss": 1.9245,
+      "step": 4141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9694691718523495e-05,
+      "loss": 1.6703,
+      "step": 4142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9694544680208307e-05,
+      "loss": 1.8798,
+      "step": 4143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9694397607043527e-05,
+      "loss": 1.9263,
+      "step": 4144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9694250499029673e-05,
+      "loss": 2.068,
+      "step": 4145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9694103356167288e-05,
+      "loss": 1.7159,
+      "step": 4146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.969395617845689e-05,
+      "loss": 1.9201,
+      "step": 4147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.969380896589901e-05,
+      "loss": 1.761,
+      "step": 4148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9693661718494187e-05,
+      "loss": 1.9838,
+      "step": 4149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.969351443624294e-05,
+      "loss": 2.1084,
+      "step": 4150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9693367119145796e-05,
+      "loss": 2.3534,
+      "step": 4151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9693219767203296e-05,
+      "loss": 1.684,
+      "step": 4152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.969307238041596e-05,
+      "loss": 1.8062,
+      "step": 4153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.969292495878432e-05,
+      "loss": 1.9453,
+      "step": 4154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9692777502308907e-05,
+      "loss": 1.7856,
+      "step": 4155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9692630010990253e-05,
+      "loss": 2.0927,
+      "step": 4156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9692482484828888e-05,
+      "loss": 2.1249,
+      "step": 4157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.969233492382534e-05,
+      "loss": 1.9941,
+      "step": 4158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9692187327980138e-05,
+      "loss": 1.8409,
+      "step": 4159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9692039697293816e-05,
+      "loss": 1.8806,
+      "step": 4160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9691892031766905e-05,
+      "loss": 2.156,
+      "step": 4161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.969174433139993e-05,
+      "loss": 1.8393,
+      "step": 4162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.969159659619343e-05,
+      "loss": 1.9926,
+      "step": 4163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.969144882614793e-05,
+      "loss": 1.9402,
+      "step": 4164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.969130102126396e-05,
+      "loss": 2.0919,
+      "step": 4165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.969115318154206e-05,
+      "loss": 1.5246,
+      "step": 4166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9691005306982757e-05,
+      "loss": 2.0176,
+      "step": 4167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9690857397586576e-05,
+      "loss": 1.762,
+      "step": 4168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9690709453354062e-05,
+      "loss": 1.5744,
+      "step": 4169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9690561474285735e-05,
+      "loss": 2.1202,
+      "step": 4170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9690413460382132e-05,
+      "loss": 2.0892,
+      "step": 4171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9690265411643783e-05,
+      "loss": 1.7039,
+      "step": 4172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9690117328071222e-05,
+      "loss": 2.0585,
+      "step": 4173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9689969209664986e-05,
+      "loss": 1.894,
+      "step": 4174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9689821056425594e-05,
+      "loss": 2.0789,
+      "step": 4175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9689672868353592e-05,
+      "loss": 1.6219,
+      "step": 4176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9689524645449508e-05,
+      "loss": 2.1768,
+      "step": 4177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.968937638771387e-05,
+      "loss": 1.9337,
+      "step": 4178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.968922809514722e-05,
+      "loss": 1.8801,
+      "step": 4179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9689079767750084e-05,
+      "loss": 2.067,
+      "step": 4180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9688931405523e-05,
+      "loss": 2.1797,
+      "step": 4181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9688783008466498e-05,
+      "loss": 1.5656,
+      "step": 4182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.968863457658111e-05,
+      "loss": 1.8544,
+      "step": 4183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9688486109867374e-05,
+      "loss": 1.906,
+      "step": 4184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.968833760832582e-05,
+      "loss": 2.0384,
+      "step": 4185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9688189071956986e-05,
+      "loss": 1.6449,
+      "step": 4186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.96880405007614e-05,
+      "loss": 1.6924,
+      "step": 4187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9687891894739605e-05,
+      "loss": 1.9098,
+      "step": 4188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9687743253892124e-05,
+      "loss": 1.7503,
+      "step": 4189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.96875945782195e-05,
+      "loss": 1.6453,
+      "step": 4190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9687445867722263e-05,
+      "loss": 1.6448,
+      "step": 4191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9687297122400952e-05,
+      "loss": 1.7548,
+      "step": 4192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9687148342256095e-05,
+      "loss": 2.09,
+      "step": 4193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9686999527288235e-05,
+      "loss": 2.0682,
+      "step": 4194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.96868506774979e-05,
+      "loss": 1.8857,
+      "step": 4195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9686701792885626e-05,
+      "loss": 1.8492,
+      "step": 4196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9686552873451953e-05,
+      "loss": 1.8856,
+      "step": 4197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9686403919197413e-05,
+      "loss": 1.7236,
+      "step": 4198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.968625493012254e-05,
+      "loss": 1.53,
+      "step": 4199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9686105906227873e-05,
+      "loss": 2.0384,
+      "step": 4200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9685956847513944e-05,
+      "loss": 1.9617,
+      "step": 4201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9685807753981292e-05,
+      "loss": 1.7849,
+      "step": 4202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.968565862563045e-05,
+      "loss": 1.8517,
+      "step": 4203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.968550946246196e-05,
+      "loss": 2.0163,
+      "step": 4204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.968536026447635e-05,
+      "loss": 1.5046,
+      "step": 4205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9685211031674163e-05,
+      "loss": 1.9212,
+      "step": 4206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9685061764055932e-05,
+      "loss": 1.7427,
+      "step": 4207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9684912461622194e-05,
+      "loss": 1.8769,
+      "step": 4208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9684763124373484e-05,
+      "loss": 1.8024,
+      "step": 4209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9684613752310344e-05,
+      "loss": 1.7467,
+      "step": 4210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.968446434543331e-05,
+      "loss": 1.8998,
+      "step": 4211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9684314903742913e-05,
+      "loss": 1.8885,
+      "step": 4212
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9684165427239693e-05,
+      "loss": 2.2349,
+      "step": 4213
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9684015915924194e-05,
+      "loss": 1.874,
+      "step": 4214
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9683866369796943e-05,
+      "loss": 1.8071,
+      "step": 4215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9683716788858484e-05,
+      "loss": 2.04,
+      "step": 4216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9683567173109355e-05,
+      "loss": 1.8025,
+      "step": 4217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.968341752255009e-05,
+      "loss": 2.1009,
+      "step": 4218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.968326783718123e-05,
+      "loss": 1.639,
+      "step": 4219
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.968311811700331e-05,
+      "loss": 1.7852,
+      "step": 4220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9682968362016874e-05,
+      "loss": 1.9932,
+      "step": 4221
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9682818572222455e-05,
+      "loss": 2.0135,
+      "step": 4222
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9682668747620593e-05,
+      "loss": 2.256,
+      "step": 4223
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.968251888821183e-05,
+      "loss": 2.1651,
+      "step": 4224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9682368993996696e-05,
+      "loss": 2.0846,
+      "step": 4225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.968221906497574e-05,
+      "loss": 2.1119,
+      "step": 4226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.968206910114949e-05,
+      "loss": 1.8513,
+      "step": 4227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9681919102518496e-05,
+      "loss": 1.8593,
+      "step": 4228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9681769069083293e-05,
+      "loss": 1.8328,
+      "step": 4229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9681619000844417e-05,
+      "loss": 1.7193,
+      "step": 4230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9681468897802412e-05,
+      "loss": 2.072,
+      "step": 4231
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9681318759957818e-05,
+      "loss": 2.0597,
+      "step": 4232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9681168587311168e-05,
+      "loss": 2.0322,
+      "step": 4233
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.968101837986301e-05,
+      "loss": 1.9719,
+      "step": 4234
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.968086813761388e-05,
+      "loss": 1.2233,
+      "step": 4235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9680717860564318e-05,
+      "loss": 2.227,
+      "step": 4236
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9680567548714864e-05,
+      "loss": 1.78,
+      "step": 4237
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.968041720206606e-05,
+      "loss": 2.0555,
+      "step": 4238
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9680266820618444e-05,
+      "loss": 2.091,
+      "step": 4239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.968011640437256e-05,
+      "loss": 2.011,
+      "step": 4240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9679965953328947e-05,
+      "loss": 1.8872,
+      "step": 4241
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9679815467488145e-05,
+      "loss": 1.9487,
+      "step": 4242
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9679664946850697e-05,
+      "loss": 2.1223,
+      "step": 4243
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.967951439141714e-05,
+      "loss": 1.8775,
+      "step": 4244
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.967936380118802e-05,
+      "loss": 1.649,
+      "step": 4245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.967921317616387e-05,
+      "loss": 1.9202,
+      "step": 4246
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9679062516345246e-05,
+      "loss": 2.0358,
+      "step": 4247
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.967891182173268e-05,
+      "loss": 2.0596,
+      "step": 4248
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9678761092326712e-05,
+      "loss": 1.3959,
+      "step": 4249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9678610328127888e-05,
+      "loss": 1.9546,
+      "step": 4250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9678459529136748e-05,
+      "loss": 2.0389,
+      "step": 4251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9678308695353835e-05,
+      "loss": 1.8734,
+      "step": 4252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.967815782677969e-05,
+      "loss": 1.8497,
+      "step": 4253
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.967800692341486e-05,
+      "loss": 1.9876,
+      "step": 4254
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.967785598525988e-05,
+      "loss": 2.1007,
+      "step": 4255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.96777050123153e-05,
+      "loss": 1.5338,
+      "step": 4256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.967755400458166e-05,
+      "loss": 1.9878,
+      "step": 4257
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9677402962059498e-05,
+      "loss": 1.9107,
+      "step": 4258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9677251884749364e-05,
+      "loss": 1.6655,
+      "step": 4259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9677100772651798e-05,
+      "loss": 1.9099,
+      "step": 4260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.967694962576734e-05,
+      "loss": 1.8833,
+      "step": 4261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9676798444096537e-05,
+      "loss": 1.6086,
+      "step": 4262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9676647227639934e-05,
+      "loss": 1.6967,
+      "step": 4263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9676495976398074e-05,
+      "loss": 1.9708,
+      "step": 4264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9676344690371497e-05,
+      "loss": 1.824,
+      "step": 4265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9676193369560753e-05,
+      "loss": 1.9389,
+      "step": 4266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9676042013966378e-05,
+      "loss": 1.7964,
+      "step": 4267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9675890623588924e-05,
+      "loss": 1.8107,
+      "step": 4268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.967573919842893e-05,
+      "loss": 1.8277,
+      "step": 4269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9675587738486935e-05,
+      "loss": 1.8826,
+      "step": 4270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.96754362437635e-05,
+      "loss": 1.8603,
+      "step": 4271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9675284714259155e-05,
+      "loss": 1.8334,
+      "step": 4272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9675133149974452e-05,
+      "loss": 1.7501,
+      "step": 4273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.967498155090993e-05,
+      "loss": 1.8869,
+      "step": 4274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.967482991706614e-05,
+      "loss": 1.932,
+      "step": 4275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9674678248443623e-05,
+      "loss": 1.8681,
+      "step": 4276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.967452654504293e-05,
+      "loss": 2.098,
+      "step": 4277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9674374806864597e-05,
+      "loss": 1.66,
+      "step": 4278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9674223033909177e-05,
+      "loss": 1.9987,
+      "step": 4279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.967407122617721e-05,
+      "loss": 1.9308,
+      "step": 4280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.967391938366925e-05,
+      "loss": 1.9354,
+      "step": 4281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.967376750638583e-05,
+      "loss": 1.6011,
+      "step": 4282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9673615594327513e-05,
+      "loss": 2.0658,
+      "step": 4283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.967346364749483e-05,
+      "loss": 1.9859,
+      "step": 4284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.967331166588833e-05,
+      "loss": 1.7402,
+      "step": 4285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9673159649508572e-05,
+      "loss": 2.0217,
+      "step": 4286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9673007598356087e-05,
+      "loss": 1.7586,
+      "step": 4287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9672855512431426e-05,
+      "loss": 1.9276,
+      "step": 4288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9672703391735137e-05,
+      "loss": 2.2079,
+      "step": 4289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9672551236267767e-05,
+      "loss": 1.8973,
+      "step": 4290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9672399046029866e-05,
+      "loss": 1.7967,
+      "step": 4291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9672246821021975e-05,
+      "loss": 1.8211,
+      "step": 4292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9672094561244645e-05,
+      "loss": 1.7803,
+      "step": 4293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9671942266698422e-05,
+      "loss": 1.7858,
+      "step": 4294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9671789937383855e-05,
+      "loss": 1.6931,
+      "step": 4295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.967163757330149e-05,
+      "loss": 1.7365,
+      "step": 4296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9671485174451872e-05,
+      "loss": 1.903,
+      "step": 4297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9671332740835555e-05,
+      "loss": 1.6916,
+      "step": 4298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9671180272453085e-05,
+      "loss": 1.8957,
+      "step": 4299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9671027769305005e-05,
+      "loss": 1.7599,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9670875231391874e-05,
+      "loss": 1.8105,
+      "step": 4301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.967072265871423e-05,
+      "loss": 1.6843,
+      "step": 4302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9670570051272624e-05,
+      "loss": 2.0279,
+      "step": 4303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9670417409067606e-05,
+      "loss": 1.6386,
+      "step": 4304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9670264732099726e-05,
+      "loss": 1.8357,
+      "step": 4305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.967011202036953e-05,
+      "loss": 1.5128,
+      "step": 4306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9669959273877567e-05,
+      "loss": 1.7322,
+      "step": 4307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.966980649262439e-05,
+      "loss": 1.8033,
+      "step": 4308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9669653676610546e-05,
+      "loss": 1.7951,
+      "step": 4309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.966950082583658e-05,
+      "loss": 1.8384,
+      "step": 4310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.966934794030305e-05,
+      "loss": 1.8507,
+      "step": 4311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9669195020010497e-05,
+      "loss": 1.8925,
+      "step": 4312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9669042064959478e-05,
+      "loss": 1.7388,
+      "step": 4313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9668889075150536e-05,
+      "loss": 1.6916,
+      "step": 4314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9668736050584224e-05,
+      "loss": 1.8282,
+      "step": 4315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9668582991261097e-05,
+      "loss": 2.2314,
+      "step": 4316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9668429897181694e-05,
+      "loss": 1.8701,
+      "step": 4317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.966827676834658e-05,
+      "loss": 2.1124,
+      "step": 4318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.966812360475629e-05,
+      "loss": 2.1741,
+      "step": 4319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.966797040641139e-05,
+      "loss": 1.6806,
+      "step": 4320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9667817173312416e-05,
+      "loss": 2.0818,
+      "step": 4321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9667663905459928e-05,
+      "loss": 1.9455,
+      "step": 4322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9667510602854475e-05,
+      "loss": 1.8674,
+      "step": 4323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9667357265496605e-05,
+      "loss": 1.6471,
+      "step": 4324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9667203893386878e-05,
+      "loss": 1.7671,
+      "step": 4325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9667050486525835e-05,
+      "loss": 1.7158,
+      "step": 4326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9666897044914032e-05,
+      "loss": 1.7026,
+      "step": 4327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9666743568552018e-05,
+      "loss": 1.6744,
+      "step": 4328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9666590057440352e-05,
+      "loss": 1.7606,
+      "step": 4329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.966643651157958e-05,
+      "loss": 1.9203,
+      "step": 4330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9666282930970252e-05,
+      "loss": 2.0849,
+      "step": 4331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9666129315612923e-05,
+      "loss": 1.6977,
+      "step": 4332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9665975665508145e-05,
+      "loss": 1.7508,
+      "step": 4333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.966582198065647e-05,
+      "loss": 1.9207,
+      "step": 4334
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9665668261058454e-05,
+      "loss": 1.9339,
+      "step": 4335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9665514506714643e-05,
+      "loss": 1.8237,
+      "step": 4336
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9665360717625595e-05,
+      "loss": 2.3149,
+      "step": 4337
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9665206893791865e-05,
+      "loss": 1.7018,
+      "step": 4338
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9665053035213996e-05,
+      "loss": 2.14,
+      "step": 4339
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.966489914189255e-05,
+      "loss": 2.0431,
+      "step": 4340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9664745213828075e-05,
+      "loss": 1.7593,
+      "step": 4341
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9664591251021128e-05,
+      "loss": 1.7984,
+      "step": 4342
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.966443725347226e-05,
+      "loss": 1.7466,
+      "step": 4343
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9664283221182025e-05,
+      "loss": 1.9329,
+      "step": 4344
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.966412915415098e-05,
+      "loss": 1.8356,
+      "step": 4345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9663975052379676e-05,
+      "loss": 1.8502,
+      "step": 4346
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9663820915868666e-05,
+      "loss": 2.0023,
+      "step": 4347
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9663666744618505e-05,
+      "loss": 2.0987,
+      "step": 4348
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9663512538629747e-05,
+      "loss": 1.9223,
+      "step": 4349
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.966335829790295e-05,
+      "loss": 2.0036,
+      "step": 4350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.966320402243866e-05,
+      "loss": 1.6897,
+      "step": 4351
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9663049712237444e-05,
+      "loss": 1.6908,
+      "step": 4352
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9662895367299843e-05,
+      "loss": 1.4084,
+      "step": 4353
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.966274098762642e-05,
+      "loss": 2.1776,
+      "step": 4354
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.966258657321773e-05,
+      "loss": 1.8327,
+      "step": 4355
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9662432124074325e-05,
+      "loss": 2.0385,
+      "step": 4356
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.966227764019676e-05,
+      "loss": 1.7143,
+      "step": 4357
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.96621231215856e-05,
+      "loss": 2.1359,
+      "step": 4358
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9661968568241382e-05,
+      "loss": 1.8431,
+      "step": 4359
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9661813980164677e-05,
+      "loss": 2.1327,
+      "step": 4360
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9661659357356034e-05,
+      "loss": 1.7956,
+      "step": 4361
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.966150469981601e-05,
+      "loss": 2.1323,
+      "step": 4362
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9661350007545167e-05,
+      "loss": 1.7418,
+      "step": 4363
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.966119528054405e-05,
+      "loss": 1.9836,
+      "step": 4364
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9661040518813223e-05,
+      "loss": 1.8632,
+      "step": 4365
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.966088572235324e-05,
+      "loss": 1.6132,
+      "step": 4366
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9660730891164656e-05,
+      "loss": 1.9272,
+      "step": 4367
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9660576025248033e-05,
+      "loss": 1.9183,
+      "step": 4368
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.966042112460392e-05,
+      "loss": 2.0164,
+      "step": 4369
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.966026618923288e-05,
+      "loss": 1.7763,
+      "step": 4370
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9660111219135465e-05,
+      "loss": 1.772,
+      "step": 4371
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9659956214312235e-05,
+      "loss": 1.8324,
+      "step": 4372
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9659801174763748e-05,
+      "loss": 2.2121,
+      "step": 4373
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9659646100490562e-05,
+      "loss": 2.0751,
+      "step": 4374
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.965949099149323e-05,
+      "loss": 1.7891,
+      "step": 4375
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9659335847772315e-05,
+      "loss": 1.8943,
+      "step": 4376
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9659180669328367e-05,
+      "loss": 1.8004,
+      "step": 4377
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9659025456161955e-05,
+      "loss": 2.0005,
+      "step": 4378
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9658870208273625e-05,
+      "loss": 1.6637,
+      "step": 4379
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9658714925663944e-05,
+      "loss": 1.8954,
+      "step": 4380
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.965855960833347e-05,
+      "loss": 1.9015,
+      "step": 4381
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9658404256282756e-05,
+      "loss": 1.987,
+      "step": 4382
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.965824886951236e-05,
+      "loss": 1.946,
+      "step": 4383
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9658093448022843e-05,
+      "loss": 1.7672,
+      "step": 4384
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.965793799181477e-05,
+      "loss": 1.7614,
+      "step": 4385
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.965778250088869e-05,
+      "loss": 1.8742,
+      "step": 4386
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9657626975245164e-05,
+      "loss": 1.7882,
+      "step": 4387
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9657471414884757e-05,
+      "loss": 1.7915,
+      "step": 4388
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9657315819808024e-05,
+      "loss": 1.4898,
+      "step": 4389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.965716019001552e-05,
+      "loss": 1.8749,
+      "step": 4390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9657004525507815e-05,
+      "loss": 1.8153,
+      "step": 4391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.965684882628546e-05,
+      "loss": 1.947,
+      "step": 4392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9656693092349016e-05,
+      "loss": 1.9335,
+      "step": 4393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9656537323699047e-05,
+      "loss": 1.8036,
+      "step": 4394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9656381520336105e-05,
+      "loss": 1.572,
+      "step": 4395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9656225682260758e-05,
+      "loss": 1.909,
+      "step": 4396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9656069809473567e-05,
+      "loss": 1.7456,
+      "step": 4397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.965591390197508e-05,
+      "loss": 2.2384,
+      "step": 4398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.965575795976587e-05,
+      "loss": 1.8208,
+      "step": 4399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9655601982846496e-05,
+      "loss": 1.6046,
+      "step": 4400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9655445971217516e-05,
+      "loss": 1.7766,
+      "step": 4401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.965528992487949e-05,
+      "loss": 1.7018,
+      "step": 4402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9655133843832983e-05,
+      "loss": 2.3904,
+      "step": 4403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.965497772807855e-05,
+      "loss": 1.7949,
+      "step": 4404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9654821577616755e-05,
+      "loss": 1.8084,
+      "step": 4405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9654665392448163e-05,
+      "loss": 1.8584,
+      "step": 4406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.965450917257333e-05,
+      "loss": 1.6697,
+      "step": 4407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9654352917992822e-05,
+      "loss": 1.807,
+      "step": 4408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.965419662870719e-05,
+      "loss": 1.7721,
+      "step": 4409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9654040304717014e-05,
+      "loss": 2.2734,
+      "step": 4410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9653883946022844e-05,
+      "loss": 1.8421,
+      "step": 4411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9653727552625242e-05,
+      "loss": 1.9105,
+      "step": 4412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.965357112452477e-05,
+      "loss": 1.8929,
+      "step": 4413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9653414661722e-05,
+      "loss": 1.9525,
+      "step": 4414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9653258164217485e-05,
+      "loss": 1.7828,
+      "step": 4415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9653101632011786e-05,
+      "loss": 2.1747,
+      "step": 4416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9652945065105472e-05,
+      "loss": 1.9505,
+      "step": 4417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9652788463499105e-05,
+      "loss": 2.1753,
+      "step": 4418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9652631827193244e-05,
+      "loss": 1.8898,
+      "step": 4419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9652475156188454e-05,
+      "loss": 2.0524,
+      "step": 4420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.96523184504853e-05,
+      "loss": 1.9025,
+      "step": 4421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9652161710084343e-05,
+      "loss": 1.9538,
+      "step": 4422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.965200493498615e-05,
+      "loss": 2.2301,
+      "step": 4423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9651848125191278e-05,
+      "loss": 1.8629,
+      "step": 4424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9651691280700297e-05,
+      "loss": 1.8761,
+      "step": 4425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.965153440151377e-05,
+      "loss": 2.048,
+      "step": 4426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9651377487632253e-05,
+      "loss": 2.4703,
+      "step": 4427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9651220539056322e-05,
+      "loss": 2.0239,
+      "step": 4428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9651063555786534e-05,
+      "loss": 1.8795,
+      "step": 4429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9650906537823453e-05,
+      "loss": 2.2098,
+      "step": 4430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9650749485167647e-05,
+      "loss": 2.0448,
+      "step": 4431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.965059239781968e-05,
+      "loss": 2.0033,
+      "step": 4432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9650435275780114e-05,
+      "loss": 1.9874,
+      "step": 4433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9650278119049515e-05,
+      "loss": 1.9729,
+      "step": 4434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9650120927628452e-05,
+      "loss": 1.8671,
+      "step": 4435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9649963701517483e-05,
+      "loss": 1.9052,
+      "step": 4436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9649806440717178e-05,
+      "loss": 2.2047,
+      "step": 4437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.96496491452281e-05,
+      "loss": 1.8936,
+      "step": 4438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9649491815050817e-05,
+      "loss": 2.1183,
+      "step": 4439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.964933445018589e-05,
+      "loss": 2.212,
+      "step": 4440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9649177050633893e-05,
+      "loss": 2.0941,
+      "step": 4441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9649019616395383e-05,
+      "loss": 1.8206,
+      "step": 4442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.964886214747093e-05,
+      "loss": 2.0585,
+      "step": 4443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9648704643861095e-05,
+      "loss": 2.1218,
+      "step": 4444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25,
+      "learning_rate": 1.9648547105566457e-05,
+      "loss": 1.5259,
+      "step": 4445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9648389532587568e-05,
+      "loss": 2.2188,
+      "step": 4446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9648231924925004e-05,
+      "loss": 1.7198,
+      "step": 4447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9648074282579324e-05,
+      "loss": 1.6764,
+      "step": 4448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9647916605551104e-05,
+      "loss": 2.3263,
+      "step": 4449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.96477588938409e-05,
+      "loss": 1.704,
+      "step": 4450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9647601147449284e-05,
+      "loss": 2.2171,
+      "step": 4451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.964744336637683e-05,
+      "loss": 1.6914,
+      "step": 4452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9647285550624094e-05,
+      "loss": 1.8098,
+      "step": 4453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.964712770019165e-05,
+      "loss": 1.7017,
+      "step": 4454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.964696981508006e-05,
+      "loss": 1.6492,
+      "step": 4455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9646811895289898e-05,
+      "loss": 1.9197,
+      "step": 4456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.964665394082173e-05,
+      "loss": 1.8044,
+      "step": 4457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9646495951676118e-05,
+      "loss": 2.1818,
+      "step": 4458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9646337927853636e-05,
+      "loss": 1.9914,
+      "step": 4459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.964617986935485e-05,
+      "loss": 1.7403,
+      "step": 4460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.964602177618033e-05,
+      "loss": 1.6756,
+      "step": 4461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9645863648330642e-05,
+      "loss": 2.1623,
+      "step": 4462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9645705485806356e-05,
+      "loss": 2.1673,
+      "step": 4463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.964554728860804e-05,
+      "loss": 1.9407,
+      "step": 4464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9645389056736262e-05,
+      "loss": 1.811,
+      "step": 4465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9645230790191593e-05,
+      "loss": 2.1586,
+      "step": 4466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.96450724889746e-05,
+      "loss": 2.097,
+      "step": 4467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.964491415308585e-05,
+      "loss": 1.9615,
+      "step": 4468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.964475578252592e-05,
+      "loss": 2.0318,
+      "step": 4469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9644597377295366e-05,
+      "loss": 1.7382,
+      "step": 4470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.964443893739477e-05,
+      "loss": 1.8252,
+      "step": 4471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9644280462824697e-05,
+      "loss": 1.9665,
+      "step": 4472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9644121953585716e-05,
+      "loss": 1.6268,
+      "step": 4473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9643963409678397e-05,
+      "loss": 1.9317,
+      "step": 4474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.964380483110331e-05,
+      "loss": 1.8365,
+      "step": 4475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9643646217861026e-05,
+      "loss": 2.0072,
+      "step": 4476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9643487569952113e-05,
+      "loss": 1.9975,
+      "step": 4477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9643328887377146e-05,
+      "loss": 2.0539,
+      "step": 4478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.964317017013669e-05,
+      "loss": 1.9706,
+      "step": 4479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9643011418231318e-05,
+      "loss": 1.8156,
+      "step": 4480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.96428526316616e-05,
+      "loss": 1.688,
+      "step": 4481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9642693810428105e-05,
+      "loss": 2.0084,
+      "step": 4482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.964253495453141e-05,
+      "loss": 2.0849,
+      "step": 4483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9642376063972083e-05,
+      "loss": 2.0059,
+      "step": 4484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.964221713875069e-05,
+      "loss": 1.7745,
+      "step": 4485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.964205817886781e-05,
+      "loss": 1.946,
+      "step": 4486
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.964189918432401e-05,
+      "loss": 1.7896,
+      "step": 4487
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9641740155119862e-05,
+      "loss": 1.8807,
+      "step": 4488
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9641581091255945e-05,
+      "loss": 2.0051,
+      "step": 4489
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9641421992732818e-05,
+      "loss": 1.8777,
+      "step": 4490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9641262859551058e-05,
+      "loss": 1.7584,
+      "step": 4491
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9641103691711237e-05,
+      "loss": 1.8352,
+      "step": 4492
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9640944489213932e-05,
+      "loss": 1.7377,
+      "step": 4493
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9640785252059714e-05,
+      "loss": 2.1014,
+      "step": 4494
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9640625980249147e-05,
+      "loss": 1.7466,
+      "step": 4495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9640466673782813e-05,
+      "loss": 1.8923,
+      "step": 4496
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9640307332661282e-05,
+      "loss": 2.1134,
+      "step": 4497
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.964014795688512e-05,
+      "loss": 1.9895,
+      "step": 4498
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9639988546454912e-05,
+      "loss": 1.7264,
+      "step": 4499
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9639829101371223e-05,
+      "loss": 1.6429,
+      "step": 4500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.963966962163463e-05,
+      "loss": 1.662,
+      "step": 4501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9639510107245702e-05,
+      "loss": 2.2942,
+      "step": 4502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9639350558205016e-05,
+      "loss": 1.8371,
+      "step": 4503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9639190974513143e-05,
+      "loss": 2.1545,
+      "step": 4504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.963903135617066e-05,
+      "loss": 1.8818,
+      "step": 4505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.963887170317814e-05,
+      "loss": 1.4284,
+      "step": 4506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9638712015536155e-05,
+      "loss": 1.854,
+      "step": 4507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9638552293245277e-05,
+      "loss": 1.7538,
+      "step": 4508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9638392536306083e-05,
+      "loss": 1.854,
+      "step": 4509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.963823274471915e-05,
+      "loss": 1.8853,
+      "step": 4510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9638072918485046e-05,
+      "loss": 2.1085,
+      "step": 4511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9637913057604355e-05,
+      "loss": 2.0004,
+      "step": 4512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9637753162077642e-05,
+      "loss": 1.7742,
+      "step": 4513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9637593231905485e-05,
+      "loss": 1.9756,
+      "step": 4514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.963743326708846e-05,
+      "loss": 1.9081,
+      "step": 4515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9637273267627143e-05,
+      "loss": 1.9668,
+      "step": 4516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9637113233522106e-05,
+      "loss": 1.7761,
+      "step": 4517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.963695316477393e-05,
+      "loss": 2.0579,
+      "step": 4518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9636793061383182e-05,
+      "loss": 2.3914,
+      "step": 4519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.963663292335044e-05,
+      "loss": 1.8735,
+      "step": 4520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9636472750676286e-05,
+      "loss": 2.158,
+      "step": 4521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9636312543361295e-05,
+      "loss": 1.9722,
+      "step": 4522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9636152301406034e-05,
+      "loss": 1.9094,
+      "step": 4523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9635992024811084e-05,
+      "loss": 2.0045,
+      "step": 4524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9635831713577022e-05,
+      "loss": 2.049,
+      "step": 4525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9635671367704426e-05,
+      "loss": 1.9704,
+      "step": 4526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.963551098719387e-05,
+      "loss": 1.889,
+      "step": 4527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.963535057204593e-05,
+      "loss": 1.7838,
+      "step": 4528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.963519012226118e-05,
+      "loss": 1.6957,
+      "step": 4529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9635029637840204e-05,
+      "loss": 1.9398,
+      "step": 4530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9634869118783576e-05,
+      "loss": 2.0466,
+      "step": 4531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.963470856509187e-05,
+      "loss": 2.0336,
+      "step": 4532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9634547976765664e-05,
+      "loss": 1.727,
+      "step": 4533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9634387353805537e-05,
+      "loss": 1.8914,
+      "step": 4534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9634226696212067e-05,
+      "loss": 1.8041,
+      "step": 4535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.963406600398583e-05,
+      "loss": 1.7619,
+      "step": 4536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9633905277127404e-05,
+      "loss": 1.8484,
+      "step": 4537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9633744515637366e-05,
+      "loss": 1.5082,
+      "step": 4538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.375,
+      "learning_rate": 1.96335837195163e-05,
+      "loss": 1.9704,
+      "step": 4539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9633422888764774e-05,
+      "loss": 1.7834,
+      "step": 4540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.963326202338337e-05,
+      "loss": 1.863,
+      "step": 4541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9633101123372673e-05,
+      "loss": 1.8567,
+      "step": 4542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.625,
+      "learning_rate": 1.963294018873325e-05,
+      "loss": 2.1833,
+      "step": 4543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9632779219465686e-05,
+      "loss": 1.9284,
+      "step": 4544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.963261821557056e-05,
+      "loss": 1.8596,
+      "step": 4545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.963245717704845e-05,
+      "loss": 1.9876,
+      "step": 4546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9632296103899935e-05,
+      "loss": 2.0951,
+      "step": 4547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.963213499612559e-05,
+      "loss": 2.1172,
+      "step": 4548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9631973853726002e-05,
+      "loss": 1.7158,
+      "step": 4549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9631812676701744e-05,
+      "loss": 1.9251,
+      "step": 4550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9631651465053397e-05,
+      "loss": 2.1052,
+      "step": 4551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.963149021878154e-05,
+      "loss": 2.19,
+      "step": 4552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9631328937886754e-05,
+      "loss": 1.9567,
+      "step": 4553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.625,
+      "learning_rate": 1.9631167622369617e-05,
+      "loss": 2.6237,
+      "step": 4554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9631006272230713e-05,
+      "loss": 2.1146,
+      "step": 4555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9630844887470617e-05,
+      "loss": 2.0496,
+      "step": 4556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9630683468089914e-05,
+      "loss": 1.6984,
+      "step": 4557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9630522014089178e-05,
+      "loss": 1.8367,
+      "step": 4558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9630360525468993e-05,
+      "loss": 1.7443,
+      "step": 4559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.963019900222994e-05,
+      "loss": 1.9397,
+      "step": 4560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.96300374443726e-05,
+      "loss": 1.7802,
+      "step": 4561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9629875851897554e-05,
+      "loss": 1.8919,
+      "step": 4562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9629714224805385e-05,
+      "loss": 1.7479,
+      "step": 4563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9629552563096663e-05,
+      "loss": 2.1419,
+      "step": 4564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.962939086677198e-05,
+      "loss": 3.1905,
+      "step": 4565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9629229135831912e-05,
+      "loss": 1.7246,
+      "step": 4566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9629067370277046e-05,
+      "loss": 1.9736,
+      "step": 4567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9628905570107958e-05,
+      "loss": 1.9297,
+      "step": 4568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9628743735325234e-05,
+      "loss": 1.8967,
+      "step": 4569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.962858186592945e-05,
+      "loss": 1.9756,
+      "step": 4570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9628419961921195e-05,
+      "loss": 1.7687,
+      "step": 4571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9628258023301044e-05,
+      "loss": 1.9287,
+      "step": 4572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9628096050069586e-05,
+      "loss": 1.9403,
+      "step": 4573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9627934042227393e-05,
+      "loss": 1.9979,
+      "step": 4574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9627771999775064e-05,
+      "loss": 1.9722,
+      "step": 4575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9627609922713162e-05,
+      "loss": 2.2104,
+      "step": 4576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9627447811042283e-05,
+      "loss": 1.9284,
+      "step": 4577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9627285664763007e-05,
+      "loss": 1.8184,
+      "step": 4578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9627123483875916e-05,
+      "loss": 1.8612,
+      "step": 4579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.962696126838159e-05,
+      "loss": 1.9248,
+      "step": 4580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9626799018280618e-05,
+      "loss": 1.75,
+      "step": 4581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.962663673357358e-05,
+      "loss": 2.5985,
+      "step": 4582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9626474414261058e-05,
+      "loss": 1.7796,
+      "step": 4583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.962631206034364e-05,
+      "loss": 1.7911,
+      "step": 4584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.96261496718219e-05,
+      "loss": 2.0604,
+      "step": 4585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9625987248696435e-05,
+      "loss": 2.2505,
+      "step": 4586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.962582479096782e-05,
+      "loss": 2.1756,
+      "step": 4587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.962566229863664e-05,
+      "loss": 1.9905,
+      "step": 4588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9625499771703483e-05,
+      "loss": 1.8677,
+      "step": 4589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.962533721016893e-05,
+      "loss": 2.0376,
+      "step": 4590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9625174614033566e-05,
+      "loss": 1.9988,
+      "step": 4591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9625011983297973e-05,
+      "loss": 1.641,
+      "step": 4592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.962484931796274e-05,
+      "loss": 2.0161,
+      "step": 4593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.962468661802845e-05,
+      "loss": 1.799,
+      "step": 4594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.962452388349569e-05,
+      "loss": 2.0258,
+      "step": 4595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.962436111436504e-05,
+      "loss": 1.8253,
+      "step": 4596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9624198310637087e-05,
+      "loss": 1.8764,
+      "step": 4597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.962403547231242e-05,
+      "loss": 1.7972,
+      "step": 4598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.962387259939162e-05,
+      "loss": 1.8949,
+      "step": 4599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9623709691875276e-05,
+      "loss": 1.7932,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.962354674976397e-05,
+      "loss": 1.9483,
+      "step": 4601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.962338377305829e-05,
+      "loss": 1.7075,
+      "step": 4602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.962322076175882e-05,
+      "loss": 2.0545,
+      "step": 4603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9623057715866146e-05,
+      "loss": 2.1131,
+      "step": 4604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.962289463538086e-05,
+      "loss": 2.2551,
+      "step": 4605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.962273152030354e-05,
+      "loss": 1.8346,
+      "step": 4606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9622568370634773e-05,
+      "loss": 1.7664,
+      "step": 4607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9622405186375154e-05,
+      "loss": 1.9562,
+      "step": 4608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.962224196752526e-05,
+      "loss": 1.8393,
+      "step": 4609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9622078714085684e-05,
+      "loss": 1.5604,
+      "step": 4610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9621915426057014e-05,
+      "loss": 1.8624,
+      "step": 4611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9621752103439827e-05,
+      "loss": 2.2307,
+      "step": 4612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9621588746234718e-05,
+      "loss": 1.8268,
+      "step": 4613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9621425354442275e-05,
+      "loss": 2.1267,
+      "step": 4614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.962126192806308e-05,
+      "loss": 1.8453,
+      "step": 4615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9621098467097727e-05,
+      "loss": 1.8378,
+      "step": 4616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.96209349715468e-05,
+      "loss": 1.9898,
+      "step": 4617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.962077144141089e-05,
+      "loss": 2.0764,
+      "step": 4618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9620607876690575e-05,
+      "loss": 2.0648,
+      "step": 4619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9620444277386454e-05,
+      "loss": 1.6549,
+      "step": 4620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.962028064349911e-05,
+      "loss": 2.1224,
+      "step": 4621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9620116975029132e-05,
+      "loss": 1.8752,
+      "step": 4622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.961995327197711e-05,
+      "loss": 1.5967,
+      "step": 4623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.961978953434363e-05,
+      "loss": 2.0351,
+      "step": 4624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.961962576212928e-05,
+      "loss": 1.7439,
+      "step": 4625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9619461955334652e-05,
+      "loss": 1.5957,
+      "step": 4626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9619298113960336e-05,
+      "loss": 2.1269,
+      "step": 4627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9619134238006917e-05,
+      "loss": 1.8387,
+      "step": 4628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9618970327474983e-05,
+      "loss": 1.7551,
+      "step": 4629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9618806382365125e-05,
+      "loss": 2.1014,
+      "step": 4630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9618642402677935e-05,
+      "loss": 1.8977,
+      "step": 4631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9618478388414e-05,
+      "loss": 1.8351,
+      "step": 4632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9618314339573906e-05,
+      "loss": 1.9412,
+      "step": 4633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.961815025615825e-05,
+      "loss": 1.8226,
+      "step": 4634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.961798613816762e-05,
+      "loss": 1.5371,
+      "step": 4635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9617821985602602e-05,
+      "loss": 1.881,
+      "step": 4636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9617657798463793e-05,
+      "loss": 1.9165,
+      "step": 4637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9617493576751775e-05,
+      "loss": 2.0989,
+      "step": 4638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9617329320467142e-05,
+      "loss": 2.0125,
+      "step": 4639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9617165029610483e-05,
+      "loss": 1.9754,
+      "step": 4640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9617000704182395e-05,
+      "loss": 1.9159,
+      "step": 4641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.961683634418346e-05,
+      "loss": 1.8668,
+      "step": 4642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.961667194961427e-05,
+      "loss": 1.642,
+      "step": 4643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9616507520475424e-05,
+      "loss": 2.2483,
+      "step": 4644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9616343056767503e-05,
+      "loss": 1.684,
+      "step": 4645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.961617855849111e-05,
+      "loss": 1.6628,
+      "step": 4646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.961601402564682e-05,
+      "loss": 2.0717,
+      "step": 4647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.961584945823524e-05,
+      "loss": 2.005,
+      "step": 4648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9615684856256955e-05,
+      "loss": 1.8488,
+      "step": 4649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9615520219712554e-05,
+      "loss": 1.7657,
+      "step": 4650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9615355548602636e-05,
+      "loss": 1.7631,
+      "step": 4651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9615190842927786e-05,
+      "loss": 1.9965,
+      "step": 4652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9615026102688598e-05,
+      "loss": 2.191,
+      "step": 4653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.961486132788567e-05,
+      "loss": 2.1448,
+      "step": 4654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9614696518519583e-05,
+      "loss": 2.3224,
+      "step": 4655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.961453167459094e-05,
+      "loss": 1.7773,
+      "step": 4656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9614366796100324e-05,
+      "loss": 1.6634,
+      "step": 4657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.961420188304834e-05,
+      "loss": 1.6484,
+      "step": 4658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.961403693543557e-05,
+      "loss": 2.0232,
+      "step": 4659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.961387195326261e-05,
+      "loss": 1.92,
+      "step": 4660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9613706936530057e-05,
+      "loss": 1.4478,
+      "step": 4661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.96135418852385e-05,
+      "loss": 1.929,
+      "step": 4662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9613376799388532e-05,
+      "loss": 2.0772,
+      "step": 4663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.961321167898075e-05,
+      "loss": 1.8131,
+      "step": 4664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9613046524015743e-05,
+      "loss": 2.1409,
+      "step": 4665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.961288133449411e-05,
+      "loss": 1.9906,
+      "step": 4666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.961271611041644e-05,
+      "loss": 1.7789,
+      "step": 4667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.961255085178333e-05,
+      "loss": 1.9542,
+      "step": 4668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9612385558595373e-05,
+      "loss": 1.8751,
+      "step": 4669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.961222023085316e-05,
+      "loss": 2.3006,
+      "step": 4670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9612054868557297e-05,
+      "loss": 1.8736,
+      "step": 4671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9611889471708363e-05,
+      "loss": 2.0944,
+      "step": 4672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.961172404030696e-05,
+      "loss": 1.6231,
+      "step": 4673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9611558574353684e-05,
+      "loss": 2.1268,
+      "step": 4674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9611393073849124e-05,
+      "loss": 1.8962,
+      "step": 4675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9611227538793883e-05,
+      "loss": 1.8348,
+      "step": 4676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9611061969188552e-05,
+      "loss": 2.1767,
+      "step": 4677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9610896365033725e-05,
+      "loss": 1.6872,
+      "step": 4678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.961073072633e-05,
+      "loss": 1.6917,
+      "step": 4679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9610565053077967e-05,
+      "loss": 1.7127,
+      "step": 4680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9610399345278227e-05,
+      "loss": 2.203,
+      "step": 4681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9610233602931375e-05,
+      "loss": 1.6395,
+      "step": 4682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.961006782603801e-05,
+      "loss": 1.9357,
+      "step": 4683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.960990201459872e-05,
+      "loss": 1.3841,
+      "step": 4684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9609736168614104e-05,
+      "loss": 1.957,
+      "step": 4685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.960957028808476e-05,
+      "loss": 1.691,
+      "step": 4686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9609404373011284e-05,
+      "loss": 1.8325,
+      "step": 4687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.960923842339427e-05,
+      "loss": 1.8463,
+      "step": 4688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9609072439234317e-05,
+      "loss": 1.8395,
+      "step": 4689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9608906420532022e-05,
+      "loss": 1.9575,
+      "step": 4690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.960874036728798e-05,
+      "loss": 1.9515,
+      "step": 4691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9608574279502793e-05,
+      "loss": 1.7831,
+      "step": 4692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.960840815717705e-05,
+      "loss": 1.6024,
+      "step": 4693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.960824200031135e-05,
+      "loss": 1.8383,
+      "step": 4694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9608075808906295e-05,
+      "loss": 2.1103,
+      "step": 4695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9607909582962478e-05,
+      "loss": 2.1524,
+      "step": 4696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.96077433224805e-05,
+      "loss": 2.0118,
+      "step": 4697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.960757702746096e-05,
+      "loss": 2.2278,
+      "step": 4698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9607410697904444e-05,
+      "loss": 1.9245,
+      "step": 4699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9607244333811566e-05,
+      "loss": 1.7485,
+      "step": 4700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9607077935182915e-05,
+      "loss": 2.0383,
+      "step": 4701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.960691150201909e-05,
+      "loss": 1.9823,
+      "step": 4702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.960674503432069e-05,
+      "loss": 1.796,
+      "step": 4703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9606578532088313e-05,
+      "loss": 1.6282,
+      "step": 4704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.960641199532256e-05,
+      "loss": 1.9521,
+      "step": 4705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9606245424024028e-05,
+      "loss": 1.9648,
+      "step": 4706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.960607881819331e-05,
+      "loss": 1.74,
+      "step": 4707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9605912177831017e-05,
+      "loss": 1.616,
+      "step": 4708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.960574550293774e-05,
+      "loss": 2.2744,
+      "step": 4709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9605578793514076e-05,
+      "loss": 1.8946,
+      "step": 4710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.960541204956063e-05,
+      "loss": 2.1022,
+      "step": 4711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9605245271078e-05,
+      "loss": 2.1456,
+      "step": 4712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9605078458066788e-05,
+      "loss": 2.2196,
+      "step": 4713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9604911610527585e-05,
+      "loss": 1.8498,
+      "step": 4714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9604744728461e-05,
+      "loss": 1.87,
+      "step": 4715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9604577811867628e-05,
+      "loss": 1.7168,
+      "step": 4716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.960441086074807e-05,
+      "loss": 1.832,
+      "step": 4717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9604243875102923e-05,
+      "loss": 1.7086,
+      "step": 4718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9604076854932795e-05,
+      "loss": 1.9821,
+      "step": 4719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9603909800238282e-05,
+      "loss": 1.8739,
+      "step": 4720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9603742711019983e-05,
+      "loss": 1.8817,
+      "step": 4721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.96035755872785e-05,
+      "loss": 2.0323,
+      "step": 4722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9603408429014434e-05,
+      "loss": 1.6987,
+      "step": 4723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9603241236228386e-05,
+      "loss": 2.0238,
+      "step": 4724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9603074008920955e-05,
+      "loss": 2.0844,
+      "step": 4725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9602906747092748e-05,
+      "loss": 1.7395,
+      "step": 4726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9602739450744363e-05,
+      "loss": 1.9137,
+      "step": 4727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9602572119876397e-05,
+      "loss": 1.8964,
+      "step": 4728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9602404754489456e-05,
+      "loss": 1.9018,
+      "step": 4729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.960223735458414e-05,
+      "loss": 1.7528,
+      "step": 4730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9602069920161054e-05,
+      "loss": 1.5978,
+      "step": 4731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9601902451220795e-05,
+      "loss": 2.0374,
+      "step": 4732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9601734947763968e-05,
+      "loss": 1.8876,
+      "step": 4733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9601567409791174e-05,
+      "loss": 1.944,
+      "step": 4734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9601399837303018e-05,
+      "loss": 1.8593,
+      "step": 4735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.96012322303001e-05,
+      "loss": 1.8225,
+      "step": 4736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.960106458878302e-05,
+      "loss": 1.582,
+      "step": 4737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9600896912752387e-05,
+      "loss": 1.9725,
+      "step": 4738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.96007292022088e-05,
+      "loss": 1.7718,
+      "step": 4739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9600561457152857e-05,
+      "loss": 2.0891,
+      "step": 4740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9600393677585172e-05,
+      "loss": 1.7745,
+      "step": 4741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.960022586350634e-05,
+      "loss": 2.2694,
+      "step": 4742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9600058014916966e-05,
+      "loss": 1.8816,
+      "step": 4743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9599890131817654e-05,
+      "loss": 1.8776,
+      "step": 4744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9599722214209006e-05,
+      "loss": 2.2345,
+      "step": 4745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9599554262091627e-05,
+      "loss": 1.6297,
+      "step": 4746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9599386275466125e-05,
+      "loss": 1.8011,
+      "step": 4747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9599218254333094e-05,
+      "loss": 1.8176,
+      "step": 4748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9599050198693147e-05,
+      "loss": 1.8519,
+      "step": 4749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9598882108546883e-05,
+      "loss": 1.6741,
+      "step": 4750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9598713983894906e-05,
+      "loss": 1.9183,
+      "step": 4751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9598545824737826e-05,
+      "loss": 1.8663,
+      "step": 4752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.959837763107624e-05,
+      "loss": 1.9922,
+      "step": 4753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.959820940291076e-05,
+      "loss": 1.8435,
+      "step": 4754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9598041140241988e-05,
+      "loss": 2.0488,
+      "step": 4755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9597872843070525e-05,
+      "loss": 1.9552,
+      "step": 4756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.959770451139698e-05,
+      "loss": 1.4962,
+      "step": 4757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9597536145221957e-05,
+      "loss": 1.9057,
+      "step": 4758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.959736774454606e-05,
+      "loss": 1.8446,
+      "step": 4759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9597199309369897e-05,
+      "loss": 1.744,
+      "step": 4760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9597030839694072e-05,
+      "loss": 1.8502,
+      "step": 4761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.959686233551919e-05,
+      "loss": 2.0137,
+      "step": 4762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.959669379684586e-05,
+      "loss": 2.7939,
+      "step": 4763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9596525223674682e-05,
+      "loss": 1.9372,
+      "step": 4764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.959635661600627e-05,
+      "loss": 2.1806,
+      "step": 4765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9596187973841218e-05,
+      "loss": 2.0994,
+      "step": 4766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9596019297180146e-05,
+      "loss": 1.9024,
+      "step": 4767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9595850586023654e-05,
+      "loss": 1.8733,
+      "step": 4768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9595681840372345e-05,
+      "loss": 1.7737,
+      "step": 4769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9595513060226832e-05,
+      "loss": 2.0044,
+      "step": 4770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9595344245587715e-05,
+      "loss": 1.8548,
+      "step": 4771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9595175396455607e-05,
+      "loss": 2.1615,
+      "step": 4772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9595006512831115e-05,
+      "loss": 1.9434,
+      "step": 4773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.959483759471484e-05,
+      "loss": 1.6048,
+      "step": 4774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9594668642107396e-05,
+      "loss": 1.6487,
+      "step": 4775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9594499655009386e-05,
+      "loss": 1.8275,
+      "step": 4776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9594330633421418e-05,
+      "loss": 1.803,
+      "step": 4777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.95941615773441e-05,
+      "loss": 1.9634,
+      "step": 4778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.959399248677804e-05,
+      "loss": 1.7549,
+      "step": 4779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.959382336172385e-05,
+      "loss": 1.8418,
+      "step": 4780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.959365420218213e-05,
+      "loss": 1.9586,
+      "step": 4781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9593485008153492e-05,
+      "loss": 1.7366,
+      "step": 4782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9593315779638546e-05,
+      "loss": 1.7468,
+      "step": 4783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9593146516637897e-05,
+      "loss": 2.1378,
+      "step": 4784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9592977219152158e-05,
+      "loss": 1.6772,
+      "step": 4785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.959280788718193e-05,
+      "loss": 1.8623,
+      "step": 4786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9592638520727826e-05,
+      "loss": 1.866,
+      "step": 4787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.959246911979046e-05,
+      "loss": 2.0035,
+      "step": 4788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9592299684370433e-05,
+      "loss": 1.8146,
+      "step": 4789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9592130214468356e-05,
+      "loss": 1.8507,
+      "step": 4790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9591960710084842e-05,
+      "loss": 2.1941,
+      "step": 4791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9591791171220497e-05,
+      "loss": 1.8013,
+      "step": 4792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.959162159787593e-05,
+      "loss": 2.0508,
+      "step": 4793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9591451990051756e-05,
+      "loss": 1.9111,
+      "step": 4794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9591282347748574e-05,
+      "loss": 2.02,
+      "step": 4795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9591112670967004e-05,
+      "loss": 1.6597,
+      "step": 4796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9590942959707652e-05,
+      "loss": 2.0858,
+      "step": 4797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.959077321397113e-05,
+      "loss": 1.7347,
+      "step": 4798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.959060343375804e-05,
+      "loss": 1.8064,
+      "step": 4799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9590433619069003e-05,
+      "loss": 2.0873,
+      "step": 4800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9590263769904627e-05,
+      "loss": 2.1681,
+      "step": 4801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.959009388626552e-05,
+      "loss": 1.9925,
+      "step": 4802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9589923968152292e-05,
+      "loss": 1.9447,
+      "step": 4803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9589754015565553e-05,
+      "loss": 1.7126,
+      "step": 4804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.958958402850592e-05,
+      "loss": 2.154,
+      "step": 4805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9589414006974e-05,
+      "loss": 1.6072,
+      "step": 4806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.95892439509704e-05,
+      "loss": 1.9176,
+      "step": 4807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.958907386049574e-05,
+      "loss": 1.8387,
+      "step": 4808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9588903735550628e-05,
+      "loss": 1.9159,
+      "step": 4809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9588733576135674e-05,
+      "loss": 2.0714,
+      "step": 4810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.958856338225149e-05,
+      "loss": 1.9111,
+      "step": 4811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9588393153898688e-05,
+      "loss": 2.5127,
+      "step": 4812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.958822289107788e-05,
+      "loss": 1.7988,
+      "step": 4813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9588052593789678e-05,
+      "loss": 1.8895,
+      "step": 4814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9587882262034697e-05,
+      "loss": 1.5348,
+      "step": 4815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9587711895813543e-05,
+      "loss": 1.8344,
+      "step": 4816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9587541495126835e-05,
+      "loss": 1.7837,
+      "step": 4817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9587371059975183e-05,
+      "loss": 1.9718,
+      "step": 4818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9587200590359197e-05,
+      "loss": 2.188,
+      "step": 4819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9587030086279494e-05,
+      "loss": 1.9271,
+      "step": 4820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9586859547736683e-05,
+      "loss": 1.6847,
+      "step": 4821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.958668897473138e-05,
+      "loss": 1.9821,
+      "step": 4822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.95865183672642e-05,
+      "loss": 2.0845,
+      "step": 4823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.958634772533575e-05,
+      "loss": 1.9052,
+      "step": 4824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.958617704894665e-05,
+      "loss": 1.9373,
+      "step": 4825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.958600633809751e-05,
+      "loss": 1.9722,
+      "step": 4826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.958583559278894e-05,
+      "loss": 1.768,
+      "step": 4827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9585664813021562e-05,
+      "loss": 1.9501,
+      "step": 4828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9585493998795984e-05,
+      "loss": 2.0595,
+      "step": 4829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9585323150112825e-05,
+      "loss": 1.5826,
+      "step": 4830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9585152266972695e-05,
+      "loss": 1.7217,
+      "step": 4831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.958498134937621e-05,
+      "loss": 2.2123,
+      "step": 4832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9584810397323982e-05,
+      "loss": 1.9555,
+      "step": 4833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9584639410816626e-05,
+      "loss": 1.8572,
+      "step": 4834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9584468389854762e-05,
+      "loss": 1.7393,
+      "step": 4835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9584297334439e-05,
+      "loss": 1.904,
+      "step": 4836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.958412624456995e-05,
+      "loss": 1.8815,
+      "step": 4837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.958395512024824e-05,
+      "loss": 2.5057,
+      "step": 4838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9583783961474475e-05,
+      "loss": 1.6912,
+      "step": 4839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9583612768249274e-05,
+      "loss": 2.0099,
+      "step": 4840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.958344154057325e-05,
+      "loss": 1.8912,
+      "step": 4841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.958327027844702e-05,
+      "loss": 2.1007,
+      "step": 4842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9583098981871204e-05,
+      "loss": 1.9867,
+      "step": 4843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9582927650846407e-05,
+      "loss": 1.8205,
+      "step": 4844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9582756285373255e-05,
+      "loss": 2.0539,
+      "step": 4845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.958258488545236e-05,
+      "loss": 1.8928,
+      "step": 4846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.958241345108434e-05,
+      "loss": 2.1629,
+      "step": 4847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9582241982269807e-05,
+      "loss": 1.8507,
+      "step": 4848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9582070479009385e-05,
+      "loss": 2.0623,
+      "step": 4849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.958189894130368e-05,
+      "loss": 1.8024,
+      "step": 4850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9581727369153316e-05,
+      "loss": 1.884,
+      "step": 4851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.958155576255891e-05,
+      "loss": 1.9863,
+      "step": 4852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9581384121521074e-05,
+      "loss": 1.9206,
+      "step": 4853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9581212446040433e-05,
+      "loss": 1.785,
+      "step": 4854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9581040736117595e-05,
+      "loss": 1.8583,
+      "step": 4855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.958086899175318e-05,
+      "loss": 1.8593,
+      "step": 4856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.958069721294781e-05,
+      "loss": 1.9221,
+      "step": 4857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.95805253997021e-05,
+      "loss": 2.0168,
+      "step": 4858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9580353552016665e-05,
+      "loss": 1.8823,
+      "step": 4859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9580181669892124e-05,
+      "loss": 1.8558,
+      "step": 4860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9580009753329098e-05,
+      "loss": 2.2393,
+      "step": 4861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9579837802328198e-05,
+      "loss": 1.6817,
+      "step": 4862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.957966581689005e-05,
+      "loss": 1.9347,
+      "step": 4863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.957949379701527e-05,
+      "loss": 1.8222,
+      "step": 4864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9579321742704477e-05,
+      "loss": 1.8367,
+      "step": 4865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.957914965395828e-05,
+      "loss": 2.1977,
+      "step": 4866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9578977530777313e-05,
+      "loss": 1.9828,
+      "step": 4867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.957880537316218e-05,
+      "loss": 1.9769,
+      "step": 4868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9578633181113513e-05,
+      "loss": 2.0881,
+      "step": 4869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9578460954631925e-05,
+      "loss": 2.2273,
+      "step": 4870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.957828869371803e-05,
+      "loss": 1.8637,
+      "step": 4871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.957811639837246e-05,
+      "loss": 1.7765,
+      "step": 4872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9577944068595823e-05,
+      "loss": 1.6858,
+      "step": 4873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9577771704388738e-05,
+      "loss": 1.885,
+      "step": 4874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.9577599305751834e-05,
+      "loss": 2.2815,
+      "step": 4875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9577426872685724e-05,
+      "loss": 2.2021,
+      "step": 4876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.957725440519103e-05,
+      "loss": 1.9329,
+      "step": 4877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.957708190326837e-05,
+      "loss": 1.9539,
+      "step": 4878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9576909366918366e-05,
+      "loss": 1.8843,
+      "step": 4879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.957673679614164e-05,
+      "loss": 1.8756,
+      "step": 4880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9576564190938807e-05,
+      "loss": 2.1403,
+      "step": 4881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9576391551310492e-05,
+      "loss": 1.8307,
+      "step": 4882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9576218877257312e-05,
+      "loss": 1.9587,
+      "step": 4883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9576046168779893e-05,
+      "loss": 1.9344,
+      "step": 4884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.957587342587885e-05,
+      "loss": 1.7512,
+      "step": 4885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9575700648554808e-05,
+      "loss": 1.6705,
+      "step": 4886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.957552783680839e-05,
+      "loss": 1.8692,
+      "step": 4887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9575354990640207e-05,
+      "loss": 2.163,
+      "step": 4888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.957518211005089e-05,
+      "loss": 1.9629,
+      "step": 4889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.957500919504106e-05,
+      "loss": 2.4937,
+      "step": 4890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9574836245611334e-05,
+      "loss": 1.6656,
+      "step": 4891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.957466326176234e-05,
+      "loss": 1.7797,
+      "step": 4892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9574490243494694e-05,
+      "loss": 1.8512,
+      "step": 4893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9574317190809018e-05,
+      "loss": 1.5413,
+      "step": 4894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9574144103705937e-05,
+      "loss": 2.2072,
+      "step": 4895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.957397098218607e-05,
+      "loss": 1.8733,
+      "step": 4896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9573797826250046e-05,
+      "loss": 1.5661,
+      "step": 4897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.957362463589848e-05,
+      "loss": 1.9107,
+      "step": 4898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9573451411131996e-05,
+      "loss": 1.762,
+      "step": 4899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.957327815195122e-05,
+      "loss": 2.1668,
+      "step": 4900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9573104858356773e-05,
+      "loss": 1.9737,
+      "step": 4901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.957293153034928e-05,
+      "loss": 1.9451,
+      "step": 4902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.957275816792936e-05,
+      "loss": 1.762,
+      "step": 4903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.957258477109764e-05,
+      "loss": 2.0581,
+      "step": 4904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9572411339854737e-05,
+      "loss": 1.9382,
+      "step": 4905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.957223787420128e-05,
+      "loss": 1.9399,
+      "step": 4906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.95720643741379e-05,
+      "loss": 1.9567,
+      "step": 4907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9571890839665205e-05,
+      "loss": 1.9813,
+      "step": 4908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9571717270783827e-05,
+      "loss": 1.3521,
+      "step": 4909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9571543667494392e-05,
+      "loss": 1.7228,
+      "step": 4910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9571370029797516e-05,
+      "loss": 2.0039,
+      "step": 4911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.957119635769383e-05,
+      "loss": 2.0439,
+      "step": 4912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.9571022651183957e-05,
+      "loss": 2.18,
+      "step": 4913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9570848910268524e-05,
+      "loss": 1.7509,
+      "step": 4914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9570675134948148e-05,
+      "loss": 1.7493,
+      "step": 4915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.957050132522346e-05,
+      "loss": 2.2799,
+      "step": 4916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9570327481095085e-05,
+      "loss": 1.8197,
+      "step": 4917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9570153602563646e-05,
+      "loss": 2.0143,
+      "step": 4918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9569979689629767e-05,
+      "loss": 1.998,
+      "step": 4919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.956980574229407e-05,
+      "loss": 2.1099,
+      "step": 4920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.956963176055719e-05,
+      "loss": 1.6907,
+      "step": 4921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9569457744419743e-05,
+      "loss": 2.0121,
+      "step": 4922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9569283693882363e-05,
+      "loss": 1.7331,
+      "step": 4923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.956910960894567e-05,
+      "loss": 1.6908,
+      "step": 4924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.956893548961029e-05,
+      "loss": 1.7195,
+      "step": 4925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.956876133587685e-05,
+      "loss": 1.8214,
+      "step": 4926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9568587147745974e-05,
+      "loss": 1.7251,
+      "step": 4927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9568412925218293e-05,
+      "loss": 1.6511,
+      "step": 4928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.956823866829443e-05,
+      "loss": 1.6636,
+      "step": 4929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9568064376975013e-05,
+      "loss": 1.6594,
+      "step": 4930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9567890051260664e-05,
+      "loss": 1.913,
+      "step": 4931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9567715691152017e-05,
+      "loss": 1.731,
+      "step": 4932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.956754129664969e-05,
+      "loss": 2.2483,
+      "step": 4933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9567366867754316e-05,
+      "loss": 1.4982,
+      "step": 4934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9567192404466522e-05,
+      "loss": 1.8745,
+      "step": 4935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9567017906786935e-05,
+      "loss": 1.4638,
+      "step": 4936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.956684337471618e-05,
+      "loss": 1.7967,
+      "step": 4937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.9566668808254884e-05,
+      "loss": 2.0969,
+      "step": 4938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9566494207403678e-05,
+      "loss": 1.6331,
+      "step": 4939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9566319572163188e-05,
+      "loss": 1.7589,
+      "step": 4940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9566144902534042e-05,
+      "loss": 2.2308,
+      "step": 4941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9565970198516865e-05,
+      "loss": 2.2868,
+      "step": 4942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.956579546011229e-05,
+      "loss": 1.8114,
+      "step": 4943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9565620687320937e-05,
+      "loss": 1.86,
+      "step": 4944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.956544588014345e-05,
+      "loss": 1.8379,
+      "step": 4945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9565271038580438e-05,
+      "loss": 1.8763,
+      "step": 4946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.956509616263254e-05,
+      "loss": 1.8581,
+      "step": 4947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.956492125230039e-05,
+      "loss": 1.7734,
+      "step": 4948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9564746307584604e-05,
+      "loss": 1.8464,
+      "step": 4949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9564571328485817e-05,
+      "loss": 1.9401,
+      "step": 4950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.956439631500466e-05,
+      "loss": 1.8422,
+      "step": 4951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.956422126714176e-05,
+      "loss": 1.6892,
+      "step": 4952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9564046184897748e-05,
+      "loss": 1.8121,
+      "step": 4953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.956387106827325e-05,
+      "loss": 2.1969,
+      "step": 4954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9563695917268896e-05,
+      "loss": 1.8361,
+      "step": 4955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9563520731885318e-05,
+      "loss": 2.0801,
+      "step": 4956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9563345512123143e-05,
+      "loss": 2.0128,
+      "step": 4957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9563170257983004e-05,
+      "loss": 2.1048,
+      "step": 4958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.956299496946553e-05,
+      "loss": 1.7129,
+      "step": 4959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9562819646571352e-05,
+      "loss": 2.1458,
+      "step": 4960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.95626442893011e-05,
+      "loss": 1.9215,
+      "step": 4961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9562468897655397e-05,
+      "loss": 1.9676,
+      "step": 4962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9562293471634886e-05,
+      "loss": 1.9752,
+      "step": 4963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9562118011240188e-05,
+      "loss": 2.1031,
+      "step": 4964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9561942516471937e-05,
+      "loss": 1.9832,
+      "step": 4965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9561766987330765e-05,
+      "loss": 1.8487,
+      "step": 4966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9561591423817304e-05,
+      "loss": 2.0348,
+      "step": 4967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.956141582593218e-05,
+      "loss": 1.9589,
+      "step": 4968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.956124019367603e-05,
+      "loss": 1.8499,
+      "step": 4969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.956106452704948e-05,
+      "loss": 1.8286,
+      "step": 4970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9560888826053162e-05,
+      "loss": 2.079,
+      "step": 4971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9560713090687718e-05,
+      "loss": 1.7869,
+      "step": 4972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9560537320953765e-05,
+      "loss": 1.8221,
+      "step": 4973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9560361516851942e-05,
+      "loss": 1.9127,
+      "step": 4974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9560185678382882e-05,
+      "loss": 1.7117,
+      "step": 4975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9560009805547215e-05,
+      "loss": 2.0904,
+      "step": 4976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9559833898345572e-05,
+      "loss": 1.8943,
+      "step": 4977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.955965795677859e-05,
+      "loss": 1.8262,
+      "step": 4978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9559481980846897e-05,
+      "loss": 1.8364,
+      "step": 4979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9559305970551125e-05,
+      "loss": 1.8136,
+      "step": 4980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.955912992589191e-05,
+      "loss": 1.9375,
+      "step": 4981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9558953846869885e-05,
+      "loss": 1.8059,
+      "step": 4982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.955877773348568e-05,
+      "loss": 1.737,
+      "step": 4983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.955860158573993e-05,
+      "loss": 1.6938,
+      "step": 4984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9558425403633267e-05,
+      "loss": 1.9865,
+      "step": 4985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.955824918716633e-05,
+      "loss": 2.0414,
+      "step": 4986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.955807293633974e-05,
+      "loss": 1.8023,
+      "step": 4987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9557896651154142e-05,
+      "loss": 2.1359,
+      "step": 4988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.955772033161017e-05,
+      "loss": 2.1835,
+      "step": 4989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9557543977708445e-05,
+      "loss": 1.9989,
+      "step": 4990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9557367589449615e-05,
+      "loss": 1.6859,
+      "step": 4991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9557191166834308e-05,
+      "loss": 2.1108,
+      "step": 4992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.955701470986316e-05,
+      "loss": 1.8435,
+      "step": 4993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9556838218536804e-05,
+      "loss": 1.9847,
+      "step": 4994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9556661692855873e-05,
+      "loss": 2.2855,
+      "step": 4995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9556485132821008e-05,
+      "loss": 1.6488,
+      "step": 4996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.955630853843283e-05,
+      "loss": 1.6046,
+      "step": 4997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.955613190969199e-05,
+      "loss": 2.0348,
+      "step": 4998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9555955246599115e-05,
+      "loss": 1.663,
+      "step": 4999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9555778549154838e-05,
+      "loss": 1.7979,
+      "step": 5000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.95556018173598e-05,
+      "loss": 1.9842,
+      "step": 5001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9555425051214635e-05,
+      "loss": 1.9872,
+      "step": 5002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.955524825071997e-05,
+      "loss": 1.8884,
+      "step": 5003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9555071415876452e-05,
+      "loss": 2.2631,
+      "step": 5004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9554894546684712e-05,
+      "loss": 1.6131,
+      "step": 5005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9554717643145385e-05,
+      "loss": 2.009,
+      "step": 5006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9554540705259107e-05,
+      "loss": 1.8253,
+      "step": 5007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9554363733026513e-05,
+      "loss": 2.0886,
+      "step": 5008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9554186726448244e-05,
+      "loss": 1.7931,
+      "step": 5009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9554009685524934e-05,
+      "loss": 1.958,
+      "step": 5010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9553832610257216e-05,
+      "loss": 1.9885,
+      "step": 5011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.955365550064573e-05,
+      "loss": 1.8676,
+      "step": 5012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9553478356691112e-05,
+      "loss": 2.0523,
+      "step": 5013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9553301178394e-05,
+      "loss": 1.8876,
+      "step": 5014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.955312396575503e-05,
+      "loss": 2.0697,
+      "step": 5015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9552946718774835e-05,
+      "loss": 2.0819,
+      "step": 5016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9552769437454057e-05,
+      "loss": 1.8948,
+      "step": 5017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9552592121793334e-05,
+      "loss": 2.1533,
+      "step": 5018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.95524147717933e-05,
+      "loss": 1.8392,
+      "step": 5019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9552237387454595e-05,
+      "loss": 2.0023,
+      "step": 5020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9552059968777853e-05,
+      "loss": 2.1707,
+      "step": 5021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9551882515763717e-05,
+      "loss": 1.6313,
+      "step": 5022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.955170502841282e-05,
+      "loss": 1.7428,
+      "step": 5023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9551527506725803e-05,
+      "loss": 1.8684,
+      "step": 5024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9551349950703303e-05,
+      "loss": 1.7313,
+      "step": 5025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9551172360345966e-05,
+      "loss": 1.8246,
+      "step": 5026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9550994735654413e-05,
+      "loss": 1.9602,
+      "step": 5027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9550817076629298e-05,
+      "loss": 1.4769,
+      "step": 5028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9550639383271254e-05,
+      "loss": 1.7743,
+      "step": 5029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9550461655580918e-05,
+      "loss": 1.7006,
+      "step": 5030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9550283893558934e-05,
+      "loss": 1.8267,
+      "step": 5031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9550106097205936e-05,
+      "loss": 2.1315,
+      "step": 5032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9549928266522563e-05,
+      "loss": 1.8318,
+      "step": 5033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.954975040150946e-05,
+      "loss": 2.2215,
+      "step": 5034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.954957250216726e-05,
+      "loss": 1.6196,
+      "step": 5035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9549394568496608e-05,
+      "loss": 1.943,
+      "step": 5036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.954921660049814e-05,
+      "loss": 2.3167,
+      "step": 5037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9549038598172495e-05,
+      "loss": 1.898,
+      "step": 5038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.954886056152032e-05,
+      "loss": 2.0197,
+      "step": 5039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9548682490542244e-05,
+      "loss": 2.0504,
+      "step": 5040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.954850438523891e-05,
+      "loss": 1.8671,
+      "step": 5041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9548326245610967e-05,
+      "loss": 1.9068,
+      "step": 5042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9548148071659046e-05,
+      "loss": 1.9861,
+      "step": 5043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9547969863383792e-05,
+      "loss": 1.7594,
+      "step": 5044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9547791620785844e-05,
+      "loss": 1.7921,
+      "step": 5045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9547613343865843e-05,
+      "loss": 2.1066,
+      "step": 5046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.954743503262443e-05,
+      "loss": 2.0722,
+      "step": 5047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.954725668706225e-05,
+      "loss": 2.0176,
+      "step": 5048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9547078307179935e-05,
+      "loss": 1.9472,
+      "step": 5049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9546899892978133e-05,
+      "loss": 1.8389,
+      "step": 5050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9546721444457484e-05,
+      "loss": 1.8731,
+      "step": 5051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9546542961618627e-05,
+      "loss": 2.008,
+      "step": 5052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9546364444462207e-05,
+      "loss": 1.9253,
+      "step": 5053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9546185892988864e-05,
+      "loss": 2.0648,
+      "step": 5054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.954600730719924e-05,
+      "loss": 1.7284,
+      "step": 5055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9545828687093982e-05,
+      "loss": 2.0657,
+      "step": 5056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.954565003267372e-05,
+      "loss": 2.2839,
+      "step": 5057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9545471343939112e-05,
+      "loss": 1.7003,
+      "step": 5058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9545292620890785e-05,
+      "loss": 1.8405,
+      "step": 5059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.954511386352939e-05,
+      "loss": 1.7753,
+      "step": 5060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9544935071855572e-05,
+      "loss": 2.137,
+      "step": 5061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9544756245869964e-05,
+      "loss": 1.8517,
+      "step": 5062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.954457738557322e-05,
+      "loss": 1.9675,
+      "step": 5063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.954439849096597e-05,
+      "loss": 1.9013,
+      "step": 5064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.954421956204887e-05,
+      "loss": 1.6133,
+      "step": 5065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9544040598822557e-05,
+      "loss": 2.0108,
+      "step": 5066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9543861601287677e-05,
+      "loss": 1.5603,
+      "step": 5067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.954368256944487e-05,
+      "loss": 1.9766,
+      "step": 5068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.954350350329478e-05,
+      "loss": 1.7925,
+      "step": 5069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.954332440283805e-05,
+      "loss": 1.9845,
+      "step": 5070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.954314526807533e-05,
+      "loss": 1.8911,
+      "step": 5071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.9542966099007258e-05,
+      "loss": 1.8597,
+      "step": 5072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9542786895634482e-05,
+      "loss": 2.0444,
+      "step": 5073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.954260765795764e-05,
+      "loss": 1.832,
+      "step": 5074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9542428385977384e-05,
+      "loss": 1.7814,
+      "step": 5075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.954224907969435e-05,
+      "loss": 1.6959,
+      "step": 5076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9542069739109192e-05,
+      "loss": 1.9189,
+      "step": 5077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9541890364222547e-05,
+      "loss": 1.6576,
+      "step": 5078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9541710955035063e-05,
+      "loss": 2.1764,
+      "step": 5079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9541531511547385e-05,
+      "loss": 2.0349,
+      "step": 5080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.954135203376016e-05,
+      "loss": 1.7591,
+      "step": 5081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9541172521674028e-05,
+      "loss": 1.8236,
+      "step": 5082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9540992975289636e-05,
+      "loss": 1.9468,
+      "step": 5083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9540813394607635e-05,
+      "loss": 1.944,
+      "step": 5084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9540633779628665e-05,
+      "loss": 2.2239,
+      "step": 5085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.954045413035337e-05,
+      "loss": 1.9972,
+      "step": 5086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.95402744467824e-05,
+      "loss": 2.0876,
+      "step": 5087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.95400947289164e-05,
+      "loss": 1.8557,
+      "step": 5088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9539914976756017e-05,
+      "loss": 2.0911,
+      "step": 5089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9539735190301894e-05,
+      "loss": 1.8522,
+      "step": 5090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.953955536955468e-05,
+      "loss": 2.2431,
+      "step": 5091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.953937551451502e-05,
+      "loss": 1.9455,
+      "step": 5092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.953919562518356e-05,
+      "loss": 1.7401,
+      "step": 5093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.953901570156095e-05,
+      "loss": 2.1242,
+      "step": 5094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9538835743647832e-05,
+      "loss": 1.9915,
+      "step": 5095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9538655751444856e-05,
+      "loss": 2.006,
+      "step": 5096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.953847572495267e-05,
+      "loss": 1.922,
+      "step": 5097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.953829566417192e-05,
+      "loss": 1.6248,
+      "step": 5098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.953811556910325e-05,
+      "loss": 1.8413,
+      "step": 5099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.953793543974731e-05,
+      "loss": 1.725,
+      "step": 5100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.953775527610475e-05,
+      "loss": 1.7724,
+      "step": 5101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9537575078176214e-05,
+      "loss": 1.827,
+      "step": 5102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9537394845962352e-05,
+      "loss": 2.1459,
+      "step": 5103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9537214579463813e-05,
+      "loss": 2.1125,
+      "step": 5104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.953703427868124e-05,
+      "loss": 1.8429,
+      "step": 5105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9536853943615287e-05,
+      "loss": 1.7429,
+      "step": 5106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9536673574266595e-05,
+      "loss": 2.052,
+      "step": 5107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.953649317063582e-05,
+      "loss": 3.3045,
+      "step": 5108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9536312732723605e-05,
+      "loss": 1.8253,
+      "step": 5109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9536132260530606e-05,
+      "loss": 1.9367,
+      "step": 5110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.953595175405746e-05,
+      "loss": 2.0002,
+      "step": 5111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.953577121330483e-05,
+      "loss": 2.1313,
+      "step": 5112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9535590638273352e-05,
+      "loss": 1.8072,
+      "step": 5113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9535410028963682e-05,
+      "loss": 1.9937,
+      "step": 5114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.953522938537647e-05,
+      "loss": 2.0805,
+      "step": 5115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9535048707512363e-05,
+      "loss": 1.9939,
+      "step": 5116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.953486799537201e-05,
+      "loss": 1.7983,
+      "step": 5117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.953468724895606e-05,
+      "loss": 1.6391,
+      "step": 5118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9534506468265167e-05,
+      "loss": 1.8068,
+      "step": 5119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9534325653299973e-05,
+      "loss": 1.706,
+      "step": 5120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.953414480406114e-05,
+      "loss": 2.0199,
+      "step": 5121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9533963920549307e-05,
+      "loss": 1.9759,
+      "step": 5122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9533783002765128e-05,
+      "loss": 2.0843,
+      "step": 5123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9533602050709258e-05,
+      "loss": 1.9115,
+      "step": 5124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.953342106438234e-05,
+      "loss": 2.0842,
+      "step": 5125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9533240043785028e-05,
+      "loss": 1.6557,
+      "step": 5126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9533058988917972e-05,
+      "loss": 1.6174,
+      "step": 5127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9532877899781827e-05,
+      "loss": 1.9351,
+      "step": 5128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9532696776377236e-05,
+      "loss": 1.8255,
+      "step": 5129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9532515618704856e-05,
+      "loss": 1.7421,
+      "step": 5130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.953233442676534e-05,
+      "loss": 1.759,
+      "step": 5131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.953215320055933e-05,
+      "loss": 1.8168,
+      "step": 5132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9531971940087485e-05,
+      "loss": 1.8551,
+      "step": 5133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.953179064535046e-05,
+      "loss": 2.0159,
+      "step": 5134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.95316093163489e-05,
+      "loss": 1.9013,
+      "step": 5135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9531427953083452e-05,
+      "loss": 1.8249,
+      "step": 5136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9531246555554784e-05,
+      "loss": 2.0492,
+      "step": 5137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9531065123763532e-05,
+      "loss": 1.8983,
+      "step": 5138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9530883657710357e-05,
+      "loss": 2.3858,
+      "step": 5139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.953070215739591e-05,
+      "loss": 1.9841,
+      "step": 5140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9530520622820843e-05,
+      "loss": 2.1751,
+      "step": 5141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9530339053985806e-05,
+      "loss": 2.0,
+      "step": 5142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9530157450891458e-05,
+      "loss": 1.9692,
+      "step": 5143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9529975813538445e-05,
+      "loss": 1.9808,
+      "step": 5144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9529794141927422e-05,
+      "loss": 2.068,
+      "step": 5145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9529612436059046e-05,
+      "loss": 1.9677,
+      "step": 5146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9529430695933964e-05,
+      "loss": 1.8669,
+      "step": 5147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9529248921552833e-05,
+      "loss": 1.9997,
+      "step": 5148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9529067112916308e-05,
+      "loss": 1.922,
+      "step": 5149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9528885270025038e-05,
+      "loss": 1.7365,
+      "step": 5150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9528703392879678e-05,
+      "loss": 1.9316,
+      "step": 5151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.952852148148088e-05,
+      "loss": 2.1372,
+      "step": 5152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.952833953582931e-05,
+      "loss": 2.4979,
+      "step": 5153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9528157555925605e-05,
+      "loss": 1.8194,
+      "step": 5154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.952797554177043e-05,
+      "loss": 1.9893,
+      "step": 5155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9527793493364434e-05,
+      "loss": 1.6609,
+      "step": 5156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9527611410708278e-05,
+      "loss": 1.7526,
+      "step": 5157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.952742929380261e-05,
+      "loss": 1.7552,
+      "step": 5158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9527247142648084e-05,
+      "loss": 1.979,
+      "step": 5159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.952706495724536e-05,
+      "loss": 1.8545,
+      "step": 5160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.9526882737595093e-05,
+      "loss": 1.9127,
+      "step": 5161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.952670048369793e-05,
+      "loss": 1.7485,
+      "step": 5162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9526518195554537e-05,
+      "loss": 1.8176,
+      "step": 5163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.952633587316556e-05,
+      "loss": 1.8708,
+      "step": 5164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.952615351653166e-05,
+      "loss": 1.8337,
+      "step": 5165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9525971125653492e-05,
+      "loss": 1.8468,
+      "step": 5166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.952578870053171e-05,
+      "loss": 1.6924,
+      "step": 5167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.952560624116697e-05,
+      "loss": 1.8774,
+      "step": 5168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.952542374755993e-05,
+      "loss": 1.862,
+      "step": 5169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9525241219711245e-05,
+      "loss": 1.9567,
+      "step": 5170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9525058657621566e-05,
+      "loss": 1.4347,
+      "step": 5171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9524876061291556e-05,
+      "loss": 1.4035,
+      "step": 5172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.952469343072187e-05,
+      "loss": 2.0103,
+      "step": 5173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9524510765913164e-05,
+      "loss": 1.8215,
+      "step": 5174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.952432806686609e-05,
+      "loss": 1.6429,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9524145333581315e-05,
+      "loss": 1.9728,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.952396256605949e-05,
+      "loss": 1.9971,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9523779764301264e-05,
+      "loss": 2.0692,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.952359692830731e-05,
+      "loss": 1.8556,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9523414058078276e-05,
+      "loss": 2.0816,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9523231153614817e-05,
+      "loss": 1.9334,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9523048214917596e-05,
+      "loss": 1.8831,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.952286524198727e-05,
+      "loss": 2.0127,
+      "step": 5183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9522682234824495e-05,
+      "loss": 2.2718,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.952249919342993e-05,
+      "loss": 2.0097,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9522316117804233e-05,
+      "loss": 1.8768,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.952213300794806e-05,
+      "loss": 1.8514,
+      "step": 5187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.952194986386207e-05,
+      "loss": 1.8313,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9521766685546923e-05,
+      "loss": 2.1071,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9521583473003276e-05,
+      "loss": 2.0741,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.952140022623179e-05,
+      "loss": 2.0314,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9521216945233118e-05,
+      "loss": 1.7162,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.952103363000793e-05,
+      "loss": 2.1022,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.952085028055687e-05,
+      "loss": 1.8301,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.952066689688061e-05,
+      "loss": 2.0672,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.95204834789798e-05,
+      "loss": 1.703,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9520300026855105e-05,
+      "loss": 2.1053,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9520116540507184e-05,
+      "loss": 1.9311,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9519933019936695e-05,
+      "loss": 1.9611,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.95197494651443e-05,
+      "loss": 1.7063,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9519565876130656e-05,
+      "loss": 1.7073,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.951938225289642e-05,
+      "loss": 2.1768,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.951919859544226e-05,
+      "loss": 1.9459,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9519014903768833e-05,
+      "loss": 1.6173,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9518831177876795e-05,
+      "loss": 2.1282,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9518647417766812e-05,
+      "loss": 1.8488,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9518463623439544e-05,
+      "loss": 2.1257,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.951827979489565e-05,
+      "loss": 2.1716,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.951809593213579e-05,
+      "loss": 1.5709,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9517912035160627e-05,
+      "loss": 2.1245,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9517728103970822e-05,
+      "loss": 1.7189,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.951754413856703e-05,
+      "loss": 1.8413,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9517360138949922e-05,
+      "loss": 1.8749,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9517176105120158e-05,
+      "loss": 1.9307,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9516992037078394e-05,
+      "loss": 2.1186,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.951680793482529e-05,
+      "loss": 2.0755,
+      "step": 5216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9516623798361518e-05,
+      "loss": 1.7748,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.951643962768773e-05,
+      "loss": 2.0893,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9516255422804595e-05,
+      "loss": 2.1312,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.951607118371277e-05,
+      "loss": 1.7071,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9515886910412922e-05,
+      "loss": 1.9148,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9515702602905708e-05,
+      "loss": 2.8698,
+      "step": 5222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9515518261191797e-05,
+      "loss": 2.0037,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9515333885271846e-05,
+      "loss": 1.7999,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9515149475146523e-05,
+      "loss": 1.9007,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9514965030816485e-05,
+      "loss": 1.7636,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.95147805522824e-05,
+      "loss": 2.1736,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9514596039544926e-05,
+      "loss": 1.97,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.951441149260473e-05,
+      "loss": 1.8754,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9514226911462476e-05,
+      "loss": 2.0254,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9514042296118823e-05,
+      "loss": 1.7037,
+      "step": 5231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9513857646574444e-05,
+      "loss": 2.0176,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9513672962829994e-05,
+      "loss": 2.1814,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.951348824488614e-05,
+      "loss": 2.0616,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.951330349274354e-05,
+      "loss": 1.7625,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9513118706402873e-05,
+      "loss": 2.0112,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9512933885864788e-05,
+      "loss": 1.9403,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.951274903112996e-05,
+      "loss": 1.672,
+      "step": 5238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9512564142199042e-05,
+      "loss": 1.781,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.951237921907271e-05,
+      "loss": 1.9532,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9512194261751625e-05,
+      "loss": 1.6816,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9512009270236453e-05,
+      "loss": 1.397,
+      "step": 5242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9511824244527852e-05,
+      "loss": 1.9324,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9511639184626496e-05,
+      "loss": 4.7686,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9511454090533045e-05,
+      "loss": 1.8795,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9511268962248168e-05,
+      "loss": 2.0371,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9511083799772524e-05,
+      "loss": 1.9798,
+      "step": 5247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9510898603106785e-05,
+      "loss": 1.9273,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9510713372251615e-05,
+      "loss": 2.0169,
+      "step": 5249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.951052810720768e-05,
+      "loss": 1.9098,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9510342807975648e-05,
+      "loss": 1.7407,
+      "step": 5251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.951015747455618e-05,
+      "loss": 1.8379,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9509972106949946e-05,
+      "loss": 1.8344,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.950978670515761e-05,
+      "loss": 1.8606,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.950960126917984e-05,
+      "loss": 1.7609,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.95094157990173e-05,
+      "loss": 1.7905,
+      "step": 5256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.950923029467067e-05,
+      "loss": 1.8977,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9509044756140592e-05,
+      "loss": 1.6904,
+      "step": 5258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9508859183427753e-05,
+      "loss": 1.9916,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9508673576532814e-05,
+      "loss": 2.0243,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9508487935456444e-05,
+      "loss": 2.1254,
+      "step": 5261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.950830226019931e-05,
+      "loss": 1.8134,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.950811655076207e-05,
+      "loss": 1.8811,
+      "step": 5263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9507930807145406e-05,
+      "loss": 2.1611,
+      "step": 5264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.950774502934998e-05,
+      "loss": 1.8316,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9507559217376454e-05,
+      "loss": 2.1657,
+      "step": 5266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9507373371225503e-05,
+      "loss": 1.7288,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9507187490897795e-05,
+      "loss": 1.6919,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9507001576393995e-05,
+      "loss": 1.7303,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9506815627714774e-05,
+      "loss": 2.057,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.95066296448608e-05,
+      "loss": 1.7502,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9506443627832738e-05,
+      "loss": 1.9857,
+      "step": 5272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9506257576631257e-05,
+      "loss": 1.825,
+      "step": 5273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9506071491257032e-05,
+      "loss": 2.0314,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.950588537171073e-05,
+      "loss": 1.9523,
+      "step": 5275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9505699217993015e-05,
+      "loss": 1.834,
+      "step": 5276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9505513030104558e-05,
+      "loss": 2.0289,
+      "step": 5277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9505326808046035e-05,
+      "loss": 1.9428,
+      "step": 5278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9505140551818106e-05,
+      "loss": 2.0183,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9504954261421443e-05,
+      "loss": 1.8162,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9504767936856722e-05,
+      "loss": 2.3012,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9504581578124606e-05,
+      "loss": 2.0281,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9504395185225765e-05,
+      "loss": 1.7293,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9504208758160875e-05,
+      "loss": 1.864,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.9504022296930596e-05,
+      "loss": 2.2834,
+      "step": 5285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9503835801535613e-05,
+      "loss": 2.003,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9503649271976582e-05,
+      "loss": 2.0088,
+      "step": 5287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.950346270825418e-05,
+      "loss": 2.1833,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9503276110369082e-05,
+      "loss": 1.8856,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9503089478321948e-05,
+      "loss": 1.7092,
+      "step": 5290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9502902812113456e-05,
+      "loss": 1.8145,
+      "step": 5291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9502716111744275e-05,
+      "loss": 2.258,
+      "step": 5292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.950252937721508e-05,
+      "loss": 2.0422,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9502342608526534e-05,
+      "loss": 2.3008,
+      "step": 5294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.950215580567932e-05,
+      "loss": 1.7691,
+      "step": 5295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.95019689686741e-05,
+      "loss": 2.1047,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.950178209751155e-05,
+      "loss": 1.9302,
+      "step": 5297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9501595192192338e-05,
+      "loss": 1.963,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9501408252717136e-05,
+      "loss": 1.7438,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9501221279086626e-05,
+      "loss": 1.8106,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9501034271301466e-05,
+      "loss": 1.6849,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9500847229362336e-05,
+      "loss": 1.8682,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.950066015326991e-05,
+      "loss": 2.086,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9500473043024856e-05,
+      "loss": 2.1514,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9500285898627848e-05,
+      "loss": 1.7292,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.950009872007956e-05,
+      "loss": 1.8986,
+      "step": 5306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.949991150738066e-05,
+      "loss": 2.0242,
+      "step": 5307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.949972426053183e-05,
+      "loss": 2.0289,
+      "step": 5308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9499536979533732e-05,
+      "loss": 1.925,
+      "step": 5309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9499349664387052e-05,
+      "loss": 1.8364,
+      "step": 5310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9499162315092452e-05,
+      "loss": 1.8229,
+      "step": 5311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.949897493165061e-05,
+      "loss": 1.9932,
+      "step": 5312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.94987875140622e-05,
+      "loss": 1.7229,
+      "step": 5313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9498600062327896e-05,
+      "loss": 1.6272,
+      "step": 5314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.949841257644837e-05,
+      "loss": 1.7785,
+      "step": 5315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.94982250564243e-05,
+      "loss": 2.0275,
+      "step": 5316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9498037502256355e-05,
+      "loss": 1.9833,
+      "step": 5317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9497849913945212e-05,
+      "loss": 1.9136,
+      "step": 5318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9497662291491543e-05,
+      "loss": 1.9439,
+      "step": 5319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.949747463489603e-05,
+      "loss": 1.7239,
+      "step": 5320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9497286944159334e-05,
+      "loss": 2.1204,
+      "step": 5321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.949709921928214e-05,
+      "loss": 1.8364,
+      "step": 5322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9496911460265123e-05,
+      "loss": 1.8676,
+      "step": 5323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9496723667108952e-05,
+      "loss": 2.0645,
+      "step": 5324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9496535839814307e-05,
+      "loss": 1.9789,
+      "step": 5325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.949634797838186e-05,
+      "loss": 1.8564,
+      "step": 5326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.949616008281229e-05,
+      "loss": 1.7313,
+      "step": 5327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.949597215310627e-05,
+      "loss": 1.9634,
+      "step": 5328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9495784189264474e-05,
+      "loss": 1.9062,
+      "step": 5329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.949559619128758e-05,
+      "loss": 1.8152,
+      "step": 5330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9495408159176266e-05,
+      "loss": 1.9445,
+      "step": 5331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.94952200929312e-05,
+      "loss": 2.1548,
+      "step": 5332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.949503199255307e-05,
+      "loss": 2.0983,
+      "step": 5333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.9494843858042542e-05,
+      "loss": 3.4087,
+      "step": 5334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9494655689400294e-05,
+      "loss": 1.7337,
+      "step": 5335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.949446748662701e-05,
+      "loss": 1.9301,
+      "step": 5336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9494279249723355e-05,
+      "loss": 1.9173,
+      "step": 5337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9494090978690012e-05,
+      "loss": 1.8404,
+      "step": 5338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.949390267352766e-05,
+      "loss": 1.7885,
+      "step": 5339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9493714334236973e-05,
+      "loss": 2.0567,
+      "step": 5340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5,
+      "learning_rate": 1.9493525960818624e-05,
+      "loss": 1.2569,
+      "step": 5341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9493337553273298e-05,
+      "loss": 1.8435,
+      "step": 5342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.949314911160167e-05,
+      "loss": 2.133,
+      "step": 5343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9492960635804413e-05,
+      "loss": 2.0061,
+      "step": 5344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.949277212588221e-05,
+      "loss": 1.7956,
+      "step": 5345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.949258358183573e-05,
+      "loss": 1.9558,
+      "step": 5346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9492395003665663e-05,
+      "loss": 1.8113,
+      "step": 5347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.949220639137268e-05,
+      "loss": 1.7872,
+      "step": 5348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.949201774495746e-05,
+      "loss": 1.8715,
+      "step": 5349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.949182906442068e-05,
+      "loss": 1.523,
+      "step": 5350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.949164034976302e-05,
+      "loss": 2.0717,
+      "step": 5351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9491451600985157e-05,
+      "loss": 2.1501,
+      "step": 5352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9491262818087772e-05,
+      "loss": 1.8381,
+      "step": 5353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.949107400107154e-05,
+      "loss": 2.0197,
+      "step": 5354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.949088514993714e-05,
+      "loss": 1.9626,
+      "step": 5355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9490696264685255e-05,
+      "loss": 1.9147,
+      "step": 5356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9490507345316557e-05,
+      "loss": 1.9389,
+      "step": 5357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9490318391831734e-05,
+      "loss": 2.049,
+      "step": 5358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9490129404231462e-05,
+      "loss": 1.5709,
+      "step": 5359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9489940382516414e-05,
+      "loss": 1.9181,
+      "step": 5360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.948975132668728e-05,
+      "loss": 1.7849,
+      "step": 5361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9489562236744733e-05,
+      "loss": 2.3049,
+      "step": 5362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9489373112689453e-05,
+      "loss": 1.9446,
+      "step": 5363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.948918395452212e-05,
+      "loss": 2.1285,
+      "step": 5364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9488994762243414e-05,
+      "loss": 2.0945,
+      "step": 5365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9488805535854018e-05,
+      "loss": 2.025,
+      "step": 5366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.948861627535461e-05,
+      "loss": 2.098,
+      "step": 5367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.948842698074587e-05,
+      "loss": 1.7345,
+      "step": 5368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.948823765202848e-05,
+      "loss": 1.7861,
+      "step": 5369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9488048289203115e-05,
+      "loss": 1.7735,
+      "step": 5370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9487858892270465e-05,
+      "loss": 1.9046,
+      "step": 5371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9487669461231203e-05,
+      "loss": 1.747,
+      "step": 5372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9487479996086018e-05,
+      "loss": 1.8333,
+      "step": 5373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.948729049683558e-05,
+      "loss": 1.7876,
+      "step": 5374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9487100963480582e-05,
+      "loss": 1.8592,
+      "step": 5375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9486911396021698e-05,
+      "loss": 1.8735,
+      "step": 5376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.948672179445961e-05,
+      "loss": 2.0363,
+      "step": 5377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9486532158795e-05,
+      "loss": 1.7115,
+      "step": 5378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.948634248902855e-05,
+      "loss": 1.6512,
+      "step": 5379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9486152785160946e-05,
+      "loss": 2.267,
+      "step": 5380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9485963047192862e-05,
+      "loss": 1.8677,
+      "step": 5381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9485773275124987e-05,
+      "loss": 2.2762,
+      "step": 5382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9485583468957998e-05,
+      "loss": 2.0007,
+      "step": 5383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9485393628692582e-05,
+      "loss": 1.9983,
+      "step": 5384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9485203754329415e-05,
+      "loss": 1.9773,
+      "step": 5385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9485013845869188e-05,
+      "loss": 2.0867,
+      "step": 5386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.948482390331258e-05,
+      "loss": 1.7174,
+      "step": 5387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9484633926660272e-05,
+      "loss": 1.9319,
+      "step": 5388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9484443915912945e-05,
+      "loss": 2.0506,
+      "step": 5389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.948425387107129e-05,
+      "loss": 2.1401,
+      "step": 5390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.948406379213598e-05,
+      "loss": 2.1555,
+      "step": 5391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9483873679107708e-05,
+      "loss": 1.9968,
+      "step": 5392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.948368353198715e-05,
+      "loss": 1.9006,
+      "step": 5393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9483493350774997e-05,
+      "loss": 1.7092,
+      "step": 5394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9483303135471924e-05,
+      "loss": 1.9884,
+      "step": 5395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9483112886078618e-05,
+      "loss": 1.7629,
+      "step": 5396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9482922602595766e-05,
+      "loss": 1.8593,
+      "step": 5397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9482732285024052e-05,
+      "loss": 1.7477,
+      "step": 5398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9482541933364155e-05,
+      "loss": 1.5584,
+      "step": 5399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9482351547616763e-05,
+      "loss": 1.9022,
+      "step": 5400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9482161127782558e-05,
+      "loss": 2.1867,
+      "step": 5401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.948197067386223e-05,
+      "loss": 2.0363,
+      "step": 5402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9481780185856456e-05,
+      "loss": 1.8618,
+      "step": 5403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9481589663765927e-05,
+      "loss": 1.7272,
+      "step": 5404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9481399107591324e-05,
+      "loss": 1.7212,
+      "step": 5405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9481208517333336e-05,
+      "loss": 1.9174,
+      "step": 5406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9481017892992645e-05,
+      "loss": 2.0507,
+      "step": 5407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9480827234569934e-05,
+      "loss": 2.136,
+      "step": 5408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9480636542065893e-05,
+      "loss": 1.8914,
+      "step": 5409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9480445815481208e-05,
+      "loss": 2.0391,
+      "step": 5410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.948025505481656e-05,
+      "loss": 1.6871,
+      "step": 5411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.948006426007264e-05,
+      "loss": 1.5595,
+      "step": 5412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9479873431250124e-05,
+      "loss": 2.3053,
+      "step": 5413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9479682568349712e-05,
+      "loss": 1.9058,
+      "step": 5414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.947949167137208e-05,
+      "loss": 1.9398,
+      "step": 5415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9479300740317917e-05,
+      "loss": 2.2387,
+      "step": 5416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.947910977518791e-05,
+      "loss": 1.8879,
+      "step": 5417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9478918775982746e-05,
+      "loss": 2.4671,
+      "step": 5418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.947872774270311e-05,
+      "loss": 1.603,
+      "step": 5419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.947853667534969e-05,
+      "loss": 1.7585,
+      "step": 5420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.947834557392317e-05,
+      "loss": 1.888,
+      "step": 5421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.947815443842424e-05,
+      "loss": 1.9338,
+      "step": 5422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9477963268853588e-05,
+      "loss": 1.7442,
+      "step": 5423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.94777720652119e-05,
+      "loss": 1.9474,
+      "step": 5424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9477580827499863e-05,
+      "loss": 2.1435,
+      "step": 5425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.947738955571816e-05,
+      "loss": 1.9896,
+      "step": 5426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9477198249867487e-05,
+      "loss": 1.8856,
+      "step": 5427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9477006909948528e-05,
+      "loss": 1.8628,
+      "step": 5428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9476815535961968e-05,
+      "loss": 2.075,
+      "step": 5429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.94766241279085e-05,
+      "loss": 1.6928,
+      "step": 5430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9476432685788807e-05,
+      "loss": 2.0362,
+      "step": 5431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.947624120960358e-05,
+      "loss": 1.8034,
+      "step": 5432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9476049699353504e-05,
+      "loss": 2.0095,
+      "step": 5433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9475858155039277e-05,
+      "loss": 2.0269,
+      "step": 5434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.947566657666158e-05,
+      "loss": 1.7315,
+      "step": 5435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9475474964221096e-05,
+      "loss": 1.8271,
+      "step": 5436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9475283317718525e-05,
+      "loss": 2.1276,
+      "step": 5437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9475091637154552e-05,
+      "loss": 1.7275,
+      "step": 5438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.947489992252986e-05,
+      "loss": 1.9565,
+      "step": 5439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.947470817384515e-05,
+      "loss": 2.1929,
+      "step": 5440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9474516391101104e-05,
+      "loss": 2.2704,
+      "step": 5441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9474324574298408e-05,
+      "loss": 1.9878,
+      "step": 5442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.947413272343776e-05,
+      "loss": 1.7544,
+      "step": 5443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9473940838519842e-05,
+      "loss": 1.8501,
+      "step": 5444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.947374891954535e-05,
+      "loss": 1.9205,
+      "step": 5445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.947355696651497e-05,
+      "loss": 1.976,
+      "step": 5446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9473364979429395e-05,
+      "loss": 2.0088,
+      "step": 5447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9473172958289314e-05,
+      "loss": 1.9237,
+      "step": 5448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9472980903095412e-05,
+      "loss": 1.8893,
+      "step": 5449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9472788813848383e-05,
+      "loss": 2.0957,
+      "step": 5450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9472596690548922e-05,
+      "loss": 1.6763,
+      "step": 5451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.947240453319772e-05,
+      "loss": 2.0115,
+      "step": 5452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.947221234179546e-05,
+      "loss": 2.0839,
+      "step": 5453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9472020116342832e-05,
+      "loss": 1.9391,
+      "step": 5454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9471827856840538e-05,
+      "loss": 1.8865,
+      "step": 5455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.947163556328926e-05,
+      "loss": 1.8538,
+      "step": 5456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9471443235689695e-05,
+      "loss": 1.8277,
+      "step": 5457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.947125087404253e-05,
+      "loss": 1.9819,
+      "step": 5458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9471058478348456e-05,
+      "loss": 2.1608,
+      "step": 5459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.947086604860817e-05,
+      "loss": 1.9741,
+      "step": 5460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9470673584822358e-05,
+      "loss": 1.687,
+      "step": 5461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9470481086991714e-05,
+      "loss": 1.7838,
+      "step": 5462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9470288555116933e-05,
+      "loss": 1.8575,
+      "step": 5463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9470095989198698e-05,
+      "loss": 2.131,
+      "step": 5464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9469903389237714e-05,
+      "loss": 2.429,
+      "step": 5465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9469710755234662e-05,
+      "loss": 1.6903,
+      "step": 5466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9469518087190244e-05,
+      "loss": 1.8002,
+      "step": 5467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9469325385105144e-05,
+      "loss": 2.0173,
+      "step": 5468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9469132648980063e-05,
+      "loss": 1.4412,
+      "step": 5469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.9468939878815686e-05,
+      "loss": 1.9623,
+      "step": 5470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9468747074612714e-05,
+      "loss": 1.962,
+      "step": 5471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.946855423637183e-05,
+      "loss": 2.009,
+      "step": 5472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9468361364093736e-05,
+      "loss": 1.8557,
+      "step": 5473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.946816845777912e-05,
+      "loss": 1.8724,
+      "step": 5474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.946797551742868e-05,
+      "loss": 2.0456,
+      "step": 5475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9467782543043103e-05,
+      "loss": 1.9534,
+      "step": 5476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.946758953462309e-05,
+      "loss": 2.0288,
+      "step": 5477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9467396492169333e-05,
+      "loss": 2.0335,
+      "step": 5478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.946720341568252e-05,
+      "loss": 2.124,
+      "step": 5479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9467010305163354e-05,
+      "loss": 1.9662,
+      "step": 5480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9466817160612522e-05,
+      "loss": 1.7397,
+      "step": 5481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9466623982030724e-05,
+      "loss": 1.9722,
+      "step": 5482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.946643076941865e-05,
+      "loss": 1.9794,
+      "step": 5483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9466237522776996e-05,
+      "loss": 1.6337,
+      "step": 5484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9466044242106455e-05,
+      "loss": 1.9098,
+      "step": 5485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9465850927407727e-05,
+      "loss": 2.1903,
+      "step": 5486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9465657578681503e-05,
+      "loss": 1.8511,
+      "step": 5487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.946546419592848e-05,
+      "loss": 1.7148,
+      "step": 5488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9465270779149348e-05,
+      "loss": 2.1824,
+      "step": 5489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9465077328344805e-05,
+      "loss": 1.8091,
+      "step": 5490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9464883843515552e-05,
+      "loss": 1.7185,
+      "step": 5491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9464690324662277e-05,
+      "loss": 1.7594,
+      "step": 5492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.946449677178568e-05,
+      "loss": 2.1708,
+      "step": 5493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9464303184886453e-05,
+      "loss": 2.0074,
+      "step": 5494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9464109563965297e-05,
+      "loss": 1.8406,
+      "step": 5495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9463915909022907e-05,
+      "loss": 1.9118,
+      "step": 5496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9463722220059974e-05,
+      "loss": 1.8499,
+      "step": 5497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9463528497077196e-05,
+      "loss": 1.686,
+      "step": 5498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9463334740075275e-05,
+      "loss": 2.0004,
+      "step": 5499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.94631409490549e-05,
+      "loss": 2.3661,
+      "step": 5500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9462947124016775e-05,
+      "loss": 1.9848,
+      "step": 5501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.946275326496159e-05,
+      "loss": 1.736,
+      "step": 5502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.946255937189005e-05,
+      "loss": 2.4027,
+      "step": 5503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.946236544480284e-05,
+      "loss": 1.744,
+      "step": 5504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.946217148370067e-05,
+      "loss": 1.8361,
+      "step": 5505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9461977488584228e-05,
+      "loss": 2.1392,
+      "step": 5506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9461783459454215e-05,
+      "loss": 1.8041,
+      "step": 5507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.946158939631133e-05,
+      "loss": 2.0419,
+      "step": 5508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9461395299156268e-05,
+      "loss": 1.9327,
+      "step": 5509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9461201167989726e-05,
+      "loss": 1.8064,
+      "step": 5510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.94610070028124e-05,
+      "loss": 1.8537,
+      "step": 5511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9460812803624997e-05,
+      "loss": 1.7258,
+      "step": 5512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.946061857042821e-05,
+      "loss": 1.9513,
+      "step": 5513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9460424303222735e-05,
+      "loss": 1.8424,
+      "step": 5514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.946023000200927e-05,
+      "loss": 1.7966,
+      "step": 5515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9460035666788516e-05,
+      "loss": 1.9348,
+      "step": 5516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9459841297561172e-05,
+      "loss": 1.6397,
+      "step": 5517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9459646894327935e-05,
+      "loss": 1.8602,
+      "step": 5518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9459452457089503e-05,
+      "loss": 1.7945,
+      "step": 5519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.945925798584658e-05,
+      "loss": 2.0029,
+      "step": 5520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9459063480599857e-05,
+      "loss": 2.0181,
+      "step": 5521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9458868941350042e-05,
+      "loss": 2.0749,
+      "step": 5522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9458674368097828e-05,
+      "loss": 1.8578,
+      "step": 5523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.945847976084392e-05,
+      "loss": 2.1375,
+      "step": 5524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.945828511958901e-05,
+      "loss": 2.0104,
+      "step": 5525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9458090444333804e-05,
+      "loss": 1.9868,
+      "step": 5526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9457895735079e-05,
+      "loss": 2.0157,
+      "step": 5527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9457700991825293e-05,
+      "loss": 1.842,
+      "step": 5528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9457506214573393e-05,
+      "loss": 1.864,
+      "step": 5529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9457311403323995e-05,
+      "loss": 2.0071,
+      "step": 5530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9457116558077795e-05,
+      "loss": 1.9646,
+      "step": 5531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.94569216788355e-05,
+      "loss": 1.7018,
+      "step": 5532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9456726765597808e-05,
+      "loss": 2.0717,
+      "step": 5533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.945653181836542e-05,
+      "loss": 1.9298,
+      "step": 5534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9456336837139037e-05,
+      "loss": 1.8142,
+      "step": 5535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.945614182191936e-05,
+      "loss": 1.8694,
+      "step": 5536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9455946772707084e-05,
+      "loss": 1.9516,
+      "step": 5537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.945575168950292e-05,
+      "loss": 1.6288,
+      "step": 5538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9455556572307562e-05,
+      "loss": 1.8402,
+      "step": 5539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.945536142112172e-05,
+      "loss": 1.8323,
+      "step": 5540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9455166235946086e-05,
+      "loss": 1.7536,
+      "step": 5541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9454971016781366e-05,
+      "loss": 1.8331,
+      "step": 5542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.945477576362826e-05,
+      "loss": 1.8581,
+      "step": 5543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.945458047648747e-05,
+      "loss": 1.9987,
+      "step": 5544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9454385155359704e-05,
+      "loss": 1.7057,
+      "step": 5545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9454189800245657e-05,
+      "loss": 2.0951,
+      "step": 5546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9453994411146032e-05,
+      "loss": 1.9886,
+      "step": 5547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9453798988061535e-05,
+      "loss": 1.8004,
+      "step": 5548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9453603530992867e-05,
+      "loss": 1.8866,
+      "step": 5549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9453408039940725e-05,
+      "loss": 1.6624,
+      "step": 5550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.945321251490582e-05,
+      "loss": 2.1236,
+      "step": 5551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9453016955888853e-05,
+      "loss": 2.1144,
+      "step": 5552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9452821362890526e-05,
+      "loss": 1.8542,
+      "step": 5553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9452625735911538e-05,
+      "loss": 1.7703,
+      "step": 5554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9452430074952603e-05,
+      "loss": 1.9009,
+      "step": 5555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.945223438001441e-05,
+      "loss": 2.1841,
+      "step": 5556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9452038651097672e-05,
+      "loss": 1.7786,
+      "step": 5557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.945184288820309e-05,
+      "loss": 2.3352,
+      "step": 5558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.945164709133137e-05,
+      "loss": 1.959,
+      "step": 5559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9451451260483215e-05,
+      "loss": 1.7973,
+      "step": 5560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9451255395659323e-05,
+      "loss": 1.7583,
+      "step": 5561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9451059496860405e-05,
+      "loss": 2.0849,
+      "step": 5562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9450863564087164e-05,
+      "loss": 2.1079,
+      "step": 5563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9450667597340306e-05,
+      "loss": 1.8961,
+      "step": 5564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9450471596620528e-05,
+      "loss": 1.8147,
+      "step": 5565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9450275561928543e-05,
+      "loss": 2.243,
+      "step": 5566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.945007949326505e-05,
+      "loss": 1.7723,
+      "step": 5567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.944988339063076e-05,
+      "loss": 2.042,
+      "step": 5568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.944968725402637e-05,
+      "loss": 1.7085,
+      "step": 5569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.944949108345259e-05,
+      "loss": 1.7364,
+      "step": 5570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9449294878910125e-05,
+      "loss": 2.1736,
+      "step": 5571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.944909864039968e-05,
+      "loss": 1.72,
+      "step": 5572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.944890236792196e-05,
+      "loss": 1.5408,
+      "step": 5573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.944870606147767e-05,
+      "loss": 1.9251,
+      "step": 5574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9448509721067516e-05,
+      "loss": 1.8894,
+      "step": 5575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9448313346692205e-05,
+      "loss": 1.9139,
+      "step": 5576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.944811693835244e-05,
+      "loss": 1.6775,
+      "step": 5577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9447920496048934e-05,
+      "loss": 2.0912,
+      "step": 5578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9447724019782382e-05,
+      "loss": 2.0559,
+      "step": 5579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.94475275095535e-05,
+      "loss": 2.0464,
+      "step": 5580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.944733096536299e-05,
+      "loss": 2.0123,
+      "step": 5581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.944713438721156e-05,
+      "loss": 1.8836,
+      "step": 5582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9446937775099914e-05,
+      "loss": 1.9374,
+      "step": 5583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.944674112902876e-05,
+      "loss": 1.6822,
+      "step": 5584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.944654444899881e-05,
+      "loss": 1.8669,
+      "step": 5585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9446347735010765e-05,
+      "loss": 2.0402,
+      "step": 5586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9446150987065335e-05,
+      "loss": 2.0287,
+      "step": 5587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9445954205163225e-05,
+      "loss": 1.7343,
+      "step": 5588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9445757389305143e-05,
+      "loss": 1.5579,
+      "step": 5589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9445560539491797e-05,
+      "loss": 1.6154,
+      "step": 5590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9445363655723896e-05,
+      "loss": 1.9563,
+      "step": 5591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9445166738002146e-05,
+      "loss": 1.9593,
+      "step": 5592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.944496978632725e-05,
+      "loss": 1.8092,
+      "step": 5593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.944477280069993e-05,
+      "loss": 2.1552,
+      "step": 5594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.944457578112088e-05,
+      "loss": 2.1445,
+      "step": 5595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9444378727590815e-05,
+      "loss": 1.947,
+      "step": 5596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9444181640110443e-05,
+      "loss": 2.0884,
+      "step": 5597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.944398451868047e-05,
+      "loss": 1.634,
+      "step": 5598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9443787363301606e-05,
+      "loss": 2.0314,
+      "step": 5599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.944359017397456e-05,
+      "loss": 2.0068,
+      "step": 5600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9443392950700045e-05,
+      "loss": 1.5607,
+      "step": 5601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.944319569347876e-05,
+      "loss": 1.9731,
+      "step": 5602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9442998402311422e-05,
+      "loss": 1.8428,
+      "step": 5603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9442801077198738e-05,
+      "loss": 1.977,
+      "step": 5604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9442603718141418e-05,
+      "loss": 1.9442,
+      "step": 5605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9442406325140165e-05,
+      "loss": 1.7624,
+      "step": 5606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9442208898195703e-05,
+      "loss": 2.073,
+      "step": 5607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9442011437308726e-05,
+      "loss": 1.8115,
+      "step": 5608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9441813942479957e-05,
+      "loss": 1.7302,
+      "step": 5609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9441616413710096e-05,
+      "loss": 1.8712,
+      "step": 5610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9441418850999856e-05,
+      "loss": 2.2555,
+      "step": 5611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.944122125434995e-05,
+      "loss": 1.8641,
+      "step": 5612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9441023623761085e-05,
+      "loss": 2.0148,
+      "step": 5613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9440825959233974e-05,
+      "loss": 1.9605,
+      "step": 5614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9440628260769326e-05,
+      "loss": 1.7724,
+      "step": 5615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9440430528367853e-05,
+      "loss": 1.9869,
+      "step": 5616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9440232762030265e-05,
+      "loss": 2.4607,
+      "step": 5617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.944003496175727e-05,
+      "loss": 1.8412,
+      "step": 5618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9439837127549587e-05,
+      "loss": 1.6183,
+      "step": 5619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9439639259407915e-05,
+      "loss": 2.1204,
+      "step": 5620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9439441357332976e-05,
+      "loss": 2.0436,
+      "step": 5621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9439243421325477e-05,
+      "loss": 2.0239,
+      "step": 5622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9439045451386132e-05,
+      "loss": 1.8752,
+      "step": 5623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.943884744751565e-05,
+      "loss": 2.0168,
+      "step": 5624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9438649409714744e-05,
+      "loss": 1.6318,
+      "step": 5625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9438451337984123e-05,
+      "loss": 1.9327,
+      "step": 5626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9438253232324503e-05,
+      "loss": 1.8014,
+      "step": 5627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9438055092736596e-05,
+      "loss": 1.891,
+      "step": 5628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9437856919221112e-05,
+      "loss": 1.9008,
+      "step": 5629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9437658711778767e-05,
+      "loss": 2.3382,
+      "step": 5630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9437460470410266e-05,
+      "loss": 1.7088,
+      "step": 5631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.943726219511633e-05,
+      "loss": 1.8147,
+      "step": 5632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9437063885897664e-05,
+      "loss": 1.7239,
+      "step": 5633
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.943686554275499e-05,
+      "loss": 2.0318,
+      "step": 5634
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9436667165689016e-05,
+      "loss": 2.0607,
+      "step": 5635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9436468754700452e-05,
+      "loss": 2.0906,
+      "step": 5636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9436270309790015e-05,
+      "loss": 1.5859,
+      "step": 5637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.943607183095842e-05,
+      "loss": 1.8661,
+      "step": 5638
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9435873318206376e-05,
+      "loss": 1.6047,
+      "step": 5639
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9435674771534603e-05,
+      "loss": 2.1363,
+      "step": 5640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9435476190943806e-05,
+      "loss": 2.2609,
+      "step": 5641
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9435277576434707e-05,
+      "loss": 1.9289,
+      "step": 5642
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9435078928008012e-05,
+      "loss": 2.1436,
+      "step": 5643
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9434880245664445e-05,
+      "loss": 2.0769,
+      "step": 5644
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9434681529404713e-05,
+      "loss": 1.5884,
+      "step": 5645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9434482779229528e-05,
+      "loss": 1.9894,
+      "step": 5646
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9434283995139614e-05,
+      "loss": 1.9604,
+      "step": 5647
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9434085177135677e-05,
+      "loss": 1.8805,
+      "step": 5648
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9433886325218438e-05,
+      "loss": 1.9141,
+      "step": 5649
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9433687439388605e-05,
+      "loss": 2.1865,
+      "step": 5650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.94334885196469e-05,
+      "loss": 1.7233,
+      "step": 5651
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.943328956599403e-05,
+      "loss": 1.6262,
+      "step": 5652
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.943309057843072e-05,
+      "loss": 2.1006,
+      "step": 5653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9432891556957682e-05,
+      "loss": 1.6899,
+      "step": 5654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9432692501575624e-05,
+      "loss": 2.2764,
+      "step": 5655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9432493412285266e-05,
+      "loss": 2.0759,
+      "step": 5656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.943229428908733e-05,
+      "loss": 1.979,
+      "step": 5657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9432095131982525e-05,
+      "loss": 2.0728,
+      "step": 5658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9431895940971567e-05,
+      "loss": 1.8776,
+      "step": 5659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9431696716055177e-05,
+      "loss": 1.7002,
+      "step": 5660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.9431497457234064e-05,
+      "loss": 1.8371,
+      "step": 5661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9431298164508953e-05,
+      "loss": 1.8096,
+      "step": 5662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9431098837880552e-05,
+      "loss": 2.242,
+      "step": 5663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9430899477349583e-05,
+      "loss": 1.569,
+      "step": 5664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9430700082916757e-05,
+      "loss": 1.8727,
+      "step": 5665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.94305006545828e-05,
+      "loss": 1.714,
+      "step": 5666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.943030119234842e-05,
+      "loss": 1.795,
+      "step": 5667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9430101696214335e-05,
+      "loss": 1.836,
+      "step": 5668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9429902166181268e-05,
+      "loss": 2.2068,
+      "step": 5669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9429702602249933e-05,
+      "loss": 1.7621,
+      "step": 5670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9429503004421044e-05,
+      "loss": 2.2505,
+      "step": 5671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9429303372695326e-05,
+      "loss": 1.8187,
+      "step": 5672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.942910370707349e-05,
+      "loss": 1.9553,
+      "step": 5673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9428904007556256e-05,
+      "loss": 2.2414,
+      "step": 5674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9428704274144345e-05,
+      "loss": 1.747,
+      "step": 5675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9428504506838467e-05,
+      "loss": 1.928,
+      "step": 5676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.942830470563935e-05,
+      "loss": 1.8367,
+      "step": 5677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9428104870547702e-05,
+      "loss": 1.8334,
+      "step": 5678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9427905001564247e-05,
+      "loss": 1.8432,
+      "step": 5679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.942770509868971e-05,
+      "loss": 2.1275,
+      "step": 5680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.94275051619248e-05,
+      "loss": 2.0543,
+      "step": 5681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.9427305191270234e-05,
+      "loss": 1.8669,
+      "step": 5682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.942710518672674e-05,
+      "loss": 1.7063,
+      "step": 5683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.942690514829503e-05,
+      "loss": 1.9994,
+      "step": 5684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9426705075975825e-05,
+      "loss": 2.2119,
+      "step": 5685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9426504969769846e-05,
+      "loss": 2.1871,
+      "step": 5686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.942630482967781e-05,
+      "loss": 2.1995,
+      "step": 5687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9426104655700437e-05,
+      "loss": 1.7066,
+      "step": 5688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.942590444783845e-05,
+      "loss": 1.85,
+      "step": 5689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9425704206092562e-05,
+      "loss": 1.8152,
+      "step": 5690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9425503930463498e-05,
+      "loss": 1.956,
+      "step": 5691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9425303620951976e-05,
+      "loss": 1.9854,
+      "step": 5692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9425103277558717e-05,
+      "loss": 1.6416,
+      "step": 5693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9424902900284438e-05,
+      "loss": 1.8352,
+      "step": 5694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9424702489129868e-05,
+      "loss": 2.0483,
+      "step": 5695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9424502044095718e-05,
+      "loss": 1.9648,
+      "step": 5696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9424301565182713e-05,
+      "loss": 1.8415,
+      "step": 5697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9424101052391568e-05,
+      "loss": 1.9574,
+      "step": 5698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9423900505723016e-05,
+      "loss": 2.1563,
+      "step": 5699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9423699925177763e-05,
+      "loss": 1.8614,
+      "step": 5700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9423499310756543e-05,
+      "loss": 1.8203,
+      "step": 5701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.942329866246007e-05,
+      "loss": 2.0536,
+      "step": 5702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9423097980289064e-05,
+      "loss": 1.9063,
+      "step": 5703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9422897264244252e-05,
+      "loss": 2.062,
+      "step": 5704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9422696514326353e-05,
+      "loss": 1.9295,
+      "step": 5705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.942249573053609e-05,
+      "loss": 1.6602,
+      "step": 5706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.942229491287418e-05,
+      "loss": 2.0959,
+      "step": 5707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.942209406134135e-05,
+      "loss": 1.8916,
+      "step": 5708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9421893175938317e-05,
+      "loss": 1.9243,
+      "step": 5709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.942169225666581e-05,
+      "loss": 2.209,
+      "step": 5710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9421491303524545e-05,
+      "loss": 1.8905,
+      "step": 5711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9421290316515246e-05,
+      "loss": 1.7767,
+      "step": 5712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.942108929563864e-05,
+      "loss": 1.8423,
+      "step": 5713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9420888240895445e-05,
+      "loss": 1.9342,
+      "step": 5714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9420687152286383e-05,
+      "loss": 1.9426,
+      "step": 5715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.942048602981218e-05,
+      "loss": 2.1279,
+      "step": 5716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9420284873473556e-05,
+      "loss": 2.064,
+      "step": 5717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9420083683271237e-05,
+      "loss": 2.0048,
+      "step": 5718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9419882459205947e-05,
+      "loss": 2.1957,
+      "step": 5719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9419681201278404e-05,
+      "loss": 1.7966,
+      "step": 5720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9419479909489337e-05,
+      "loss": 2.2518,
+      "step": 5721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9419278583839464e-05,
+      "loss": 1.5584,
+      "step": 5722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9419077224329516e-05,
+      "loss": 1.6643,
+      "step": 5723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9418875830960212e-05,
+      "loss": 1.5185,
+      "step": 5724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9418674403732277e-05,
+      "loss": 1.9617,
+      "step": 5725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.941847294264643e-05,
+      "loss": 1.9163,
+      "step": 5726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.941827144770341e-05,
+      "loss": 1.7391,
+      "step": 5727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9418069918903925e-05,
+      "loss": 2.1211,
+      "step": 5728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9417868356248703e-05,
+      "loss": 2.0006,
+      "step": 5729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9417666759738476e-05,
+      "loss": 1.931,
+      "step": 5730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9417465129373965e-05,
+      "loss": 1.9106,
+      "step": 5731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.941726346515589e-05,
+      "loss": 2.0766,
+      "step": 5732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9417061767084984e-05,
+      "loss": 1.8175,
+      "step": 5733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9416860035161965e-05,
+      "loss": 2.0269,
+      "step": 5734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.941665826938757e-05,
+      "loss": 2.0816,
+      "step": 5735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9416456469762503e-05,
+      "loss": 1.9883,
+      "step": 5736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9416254636287508e-05,
+      "loss": 2.2742,
+      "step": 5737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9416052768963304e-05,
+      "loss": 1.7705,
+      "step": 5738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.941585086779062e-05,
+      "loss": 1.9186,
+      "step": 5739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9415648932770173e-05,
+      "loss": 1.8054,
+      "step": 5740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9415446963902694e-05,
+      "loss": 2.0379,
+      "step": 5741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9415244961188916e-05,
+      "loss": 1.9354,
+      "step": 5742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9415042924629555e-05,
+      "loss": 1.9047,
+      "step": 5743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9414840854225347e-05,
+      "loss": 1.7951,
+      "step": 5744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9414638749977005e-05,
+      "loss": 1.882,
+      "step": 5745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9414436611885268e-05,
+      "loss": 1.8333,
+      "step": 5746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9414234439950855e-05,
+      "loss": 1.894,
+      "step": 5747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9414032234174498e-05,
+      "loss": 1.9306,
+      "step": 5748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.941382999455692e-05,
+      "loss": 1.7965,
+      "step": 5749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.941362772109885e-05,
+      "loss": 2.0399,
+      "step": 5750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9413425413801014e-05,
+      "loss": 1.7384,
+      "step": 5751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9413223072664143e-05,
+      "loss": 2.0706,
+      "step": 5752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9413020697688957e-05,
+      "loss": 2.1478,
+      "step": 5753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9412818288876186e-05,
+      "loss": 2.181,
+      "step": 5754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9412615846226567e-05,
+      "loss": 1.7976,
+      "step": 5755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9412413369740815e-05,
+      "loss": 2.0348,
+      "step": 5756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9412210859419663e-05,
+      "loss": 1.5815,
+      "step": 5757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.941200831526384e-05,
+      "loss": 2.1357,
+      "step": 5758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9411805737274072e-05,
+      "loss": 1.8863,
+      "step": 5759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.941160312545109e-05,
+      "loss": 1.8103,
+      "step": 5760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9411400479795618e-05,
+      "loss": 1.8789,
+      "step": 5761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9411197800308387e-05,
+      "loss": 1.7237,
+      "step": 5762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.941099508699013e-05,
+      "loss": 1.958,
+      "step": 5763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9410792339841564e-05,
+      "loss": 1.7135,
+      "step": 5764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.941058955886343e-05,
+      "loss": 2.1936,
+      "step": 5765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.941038674405645e-05,
+      "loss": 1.5707,
+      "step": 5766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9410183895421357e-05,
+      "loss": 1.7643,
+      "step": 5767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.940998101295888e-05,
+      "loss": 1.7029,
+      "step": 5768
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9409778096669744e-05,
+      "loss": 2.2365,
+      "step": 5769
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9409575146554684e-05,
+      "loss": 1.9456,
+      "step": 5770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9409372162614423e-05,
+      "loss": 1.7615,
+      "step": 5771
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9409169144849694e-05,
+      "loss": 1.9132,
+      "step": 5772
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.940896609326123e-05,
+      "loss": 1.7157,
+      "step": 5773
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9408763007849754e-05,
+      "loss": 1.8644,
+      "step": 5774
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9408559888616004e-05,
+      "loss": 1.8434,
+      "step": 5775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9408356735560706e-05,
+      "loss": 1.784,
+      "step": 5776
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.940815354868459e-05,
+      "loss": 2.1456,
+      "step": 5777
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9407950327988384e-05,
+      "loss": 1.7327,
+      "step": 5778
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9407747073472826e-05,
+      "loss": 2.0604,
+      "step": 5779
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.940754378513864e-05,
+      "loss": 2.1908,
+      "step": 5780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9407340462986558e-05,
+      "loss": 1.9123,
+      "step": 5781
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9407137107017314e-05,
+      "loss": 1.7118,
+      "step": 5782
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9406933717231635e-05,
+      "loss": 1.9724,
+      "step": 5783
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9406730293630253e-05,
+      "loss": 1.705,
+      "step": 5784
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9406526836213903e-05,
+      "loss": 1.9847,
+      "step": 5785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9406323344983313e-05,
+      "loss": 2.0107,
+      "step": 5786
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9406119819939215e-05,
+      "loss": 2.1073,
+      "step": 5787
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.940591626108234e-05,
+      "loss": 2.4726,
+      "step": 5788
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.940571266841342e-05,
+      "loss": 1.9033,
+      "step": 5789
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.940550904193319e-05,
+      "loss": 2.4098,
+      "step": 5790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9405305381642376e-05,
+      "loss": 1.5228,
+      "step": 5791
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9405101687541716e-05,
+      "loss": 1.9284,
+      "step": 5792
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.940489795963194e-05,
+      "loss": 1.6142,
+      "step": 5793
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9404694197913776e-05,
+      "loss": 1.9547,
+      "step": 5794
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9404490402387965e-05,
+      "loss": 1.8254,
+      "step": 5795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9404286573055233e-05,
+      "loss": 1.9527,
+      "step": 5796
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9404082709916318e-05,
+      "loss": 2.0473,
+      "step": 5797
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9403878812971942e-05,
+      "loss": 1.9705,
+      "step": 5798
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.940367488222285e-05,
+      "loss": 1.8123,
+      "step": 5799
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9403470917669772e-05,
+      "loss": 2.324,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.940326691931344e-05,
+      "loss": 1.819,
+      "step": 5801
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9403062887154587e-05,
+      "loss": 2.1244,
+      "step": 5802
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9402858821193945e-05,
+      "loss": 1.8047,
+      "step": 5803
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.940265472143225e-05,
+      "loss": 2.0607,
+      "step": 5804
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9402450587870233e-05,
+      "loss": 1.7079,
+      "step": 5805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9402246420508634e-05,
+      "loss": 1.9088,
+      "step": 5806
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.940204221934818e-05,
+      "loss": 1.9237,
+      "step": 5807
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9401837984389608e-05,
+      "loss": 1.8804,
+      "step": 5808
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9401633715633652e-05,
+      "loss": 1.8163,
+      "step": 5809
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9401429413081048e-05,
+      "loss": 1.7126,
+      "step": 5810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9401225076732527e-05,
+      "loss": 1.9683,
+      "step": 5811
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9401020706588824e-05,
+      "loss": 2.0446,
+      "step": 5812
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9400816302650675e-05,
+      "loss": 1.8073,
+      "step": 5813
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9400611864918816e-05,
+      "loss": 1.7883,
+      "step": 5814
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.940040739339398e-05,
+      "loss": 2.0282,
+      "step": 5815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9400202888076904e-05,
+      "loss": 1.716,
+      "step": 5816
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.939999834896832e-05,
+      "loss": 2.0832,
+      "step": 5817
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.939979377606897e-05,
+      "loss": 1.7891,
+      "step": 5818
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9399589169379576e-05,
+      "loss": 1.548,
+      "step": 5819
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9399384528900886e-05,
+      "loss": 1.993,
+      "step": 5820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9399179854633633e-05,
+      "loss": 1.9754,
+      "step": 5821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.939897514657855e-05,
+      "loss": 2.1402,
+      "step": 5822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.939877040473637e-05,
+      "loss": 2.0492,
+      "step": 5823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9398565629107838e-05,
+      "loss": 1.8282,
+      "step": 5824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9398360819693686e-05,
+      "loss": 1.703,
+      "step": 5825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9398155976494646e-05,
+      "loss": 1.8178,
+      "step": 5826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.939795109951146e-05,
+      "loss": 1.7697,
+      "step": 5827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.939774618874486e-05,
+      "loss": 1.8447,
+      "step": 5828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9397541244195588e-05,
+      "loss": 2.1296,
+      "step": 5829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9397336265864374e-05,
+      "loss": 1.92,
+      "step": 5830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9397131253751964e-05,
+      "loss": 2.0447,
+      "step": 5831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 2.0309,
+      "step": 5832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.939672112818648e-05,
+      "loss": 2.1084,
+      "step": 5833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9396516014734884e-05,
+      "loss": 1.8568,
+      "step": 5834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9396310867505036e-05,
+      "loss": 1.6912,
+      "step": 5835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9396105686497675e-05,
+      "loss": 2.0693,
+      "step": 5836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.939590047171353e-05,
+      "loss": 1.7148,
+      "step": 5837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9395695223153352e-05,
+      "loss": 2.0931,
+      "step": 5838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9395489940817866e-05,
+      "loss": 1.7708,
+      "step": 5839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.939528462470782e-05,
+      "loss": 2.6576,
+      "step": 5840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9395079274823945e-05,
+      "loss": 2.1132,
+      "step": 5841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9394873891166985e-05,
+      "loss": 1.7556,
+      "step": 5842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.939466847373767e-05,
+      "loss": 2.2803,
+      "step": 5843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9394463022536746e-05,
+      "loss": 1.9968,
+      "step": 5844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.939425753756495e-05,
+      "loss": 1.9678,
+      "step": 5845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.939405201882302e-05,
+      "loss": 2.0437,
+      "step": 5846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.939384646631169e-05,
+      "loss": 1.8831,
+      "step": 5847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.939364088003171e-05,
+      "loss": 1.9542,
+      "step": 5848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.939343525998381e-05,
+      "loss": 2.0096,
+      "step": 5849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9393229606168732e-05,
+      "loss": 1.6248,
+      "step": 5850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9393023918587216e-05,
+      "loss": 1.6697,
+      "step": 5851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9392818197239998e-05,
+      "loss": 1.7813,
+      "step": 5852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.939261244212782e-05,
+      "loss": 1.9169,
+      "step": 5853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9392406653251423e-05,
+      "loss": 1.75,
+      "step": 5854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9392200830611544e-05,
+      "loss": 1.6202,
+      "step": 5855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9391994974208924e-05,
+      "loss": 1.8387,
+      "step": 5856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9391789084044306e-05,
+      "loss": 2.087,
+      "step": 5857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9391583160118425e-05,
+      "loss": 1.9634,
+      "step": 5858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9391377202432027e-05,
+      "loss": 1.7632,
+      "step": 5859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9391171210985847e-05,
+      "loss": 2.5524,
+      "step": 5860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9390965185780626e-05,
+      "loss": 1.8572,
+      "step": 5861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9390759126817107e-05,
+      "loss": 2.0591,
+      "step": 5862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9390553034096026e-05,
+      "loss": 1.9006,
+      "step": 5863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9390346907618133e-05,
+      "loss": 1.8329,
+      "step": 5864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.939014074738416e-05,
+      "loss": 1.7593,
+      "step": 5865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9389934553394856e-05,
+      "loss": 2.1088,
+      "step": 5866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9389728325650953e-05,
+      "loss": 1.9695,
+      "step": 5867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.93895220641532e-05,
+      "loss": 1.9818,
+      "step": 5868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9389315768902335e-05,
+      "loss": 1.9315,
+      "step": 5869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9389109439899102e-05,
+      "loss": 1.8782,
+      "step": 5870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9388903077144242e-05,
+      "loss": 1.7679,
+      "step": 5871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9388696680638494e-05,
+      "loss": 1.8311,
+      "step": 5872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.93884902503826e-05,
+      "loss": 2.146,
+      "step": 5873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9388283786377306e-05,
+      "loss": 1.5981,
+      "step": 5874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.938807728862335e-05,
+      "loss": 1.91,
+      "step": 5875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.938787075712148e-05,
+      "loss": 1.7192,
+      "step": 5876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.938766419187243e-05,
+      "loss": 1.9165,
+      "step": 5877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.938745759287695e-05,
+      "loss": 1.8321,
+      "step": 5878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.938725096013578e-05,
+      "loss": 2.3752,
+      "step": 5879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9387044293649663e-05,
+      "loss": 1.954,
+      "step": 5880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.938683759341934e-05,
+      "loss": 1.9299,
+      "step": 5881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9386630859445558e-05,
+      "loss": 1.9059,
+      "step": 5882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9386424091729057e-05,
+      "loss": 1.8229,
+      "step": 5883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9386217290270582e-05,
+      "loss": 1.55,
+      "step": 5884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9386010455070875e-05,
+      "loss": 2.2822,
+      "step": 5885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.938580358613068e-05,
+      "loss": 1.8524,
+      "step": 5886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.938559668345074e-05,
+      "loss": 1.6255,
+      "step": 5887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9385389747031803e-05,
+      "loss": 1.6696,
+      "step": 5888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9385182776874604e-05,
+      "loss": 2.0677,
+      "step": 5889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9384975772979896e-05,
+      "loss": 1.8031,
+      "step": 5890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.938476873534842e-05,
+      "loss": 1.9231,
+      "step": 5891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.938456166398092e-05,
+      "loss": 2.2937,
+      "step": 5892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9384354558878137e-05,
+      "loss": 1.8768,
+      "step": 5893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.938414742004082e-05,
+      "loss": 2.045,
+      "step": 5894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9383940247469714e-05,
+      "loss": 1.8681,
+      "step": 5895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9383733041165563e-05,
+      "loss": 2.4345,
+      "step": 5896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9383525801129108e-05,
+      "loss": 1.8855,
+      "step": 5897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.9383318527361097e-05,
+      "loss": 3.0944,
+      "step": 5898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9383111219862276e-05,
+      "loss": 2.0679,
+      "step": 5899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9382903878633387e-05,
+      "loss": 1.5118,
+      "step": 5900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.938269650367518e-05,
+      "loss": 1.8228,
+      "step": 5901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.93824890949884e-05,
+      "loss": 1.7125,
+      "step": 5902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9382281652573787e-05,
+      "loss": 1.9911,
+      "step": 5903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.938207417643209e-05,
+      "loss": 1.7547,
+      "step": 5904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.938186666656406e-05,
+      "loss": 1.9796,
+      "step": 5905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9381659122970433e-05,
+      "loss": 1.8748,
+      "step": 5906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9381451545651964e-05,
+      "loss": 2.0233,
+      "step": 5907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9381243934609392e-05,
+      "loss": 1.9747,
+      "step": 5908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9381036289843468e-05,
+      "loss": 1.9755,
+      "step": 5909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9380828611354936e-05,
+      "loss": 2.495,
+      "step": 5910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9380620899144545e-05,
+      "loss": 1.9203,
+      "step": 5911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.938041315321304e-05,
+      "loss": 2.1101,
+      "step": 5912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.938020537356117e-05,
+      "loss": 1.9359,
+      "step": 5913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9379997560189677e-05,
+      "loss": 1.8517,
+      "step": 5914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9379789713099312e-05,
+      "loss": 1.6461,
+      "step": 5915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.937958183229082e-05,
+      "loss": 1.892,
+      "step": 5916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.937937391776495e-05,
+      "loss": 1.7615,
+      "step": 5917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.937916596952245e-05,
+      "loss": 1.9297,
+      "step": 5918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9378957987564066e-05,
+      "loss": 1.8607,
+      "step": 5919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.937874997189055e-05,
+      "loss": 1.738,
+      "step": 5920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.937854192250264e-05,
+      "loss": 1.9682,
+      "step": 5921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.937833383940109e-05,
+      "loss": 1.7937,
+      "step": 5922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.937812572258665e-05,
+      "loss": 2.0935,
+      "step": 5923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.937791757206007e-05,
+      "loss": 1.9749,
+      "step": 5924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.937770938782209e-05,
+      "loss": 1.9007,
+      "step": 5925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9377501169873462e-05,
+      "loss": 1.8277,
+      "step": 5926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9377292918214933e-05,
+      "loss": 1.569,
+      "step": 5927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9377084632847257e-05,
+      "loss": 2.2059,
+      "step": 5928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9376876313771177e-05,
+      "loss": 1.7742,
+      "step": 5929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9376667960987444e-05,
+      "loss": 1.8911,
+      "step": 5930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.937645957449681e-05,
+      "loss": 2.2229,
+      "step": 5931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.937625115430002e-05,
+      "loss": 1.8595,
+      "step": 5932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9376042700397826e-05,
+      "loss": 1.8914,
+      "step": 5933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9375834212790975e-05,
+      "loss": 1.5864,
+      "step": 5934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9375625691480215e-05,
+      "loss": 2.0216,
+      "step": 5935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.93754171364663e-05,
+      "loss": 1.8542,
+      "step": 5936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.937520854774998e-05,
+      "loss": 2.0369,
+      "step": 5937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9374999925331996e-05,
+      "loss": 1.7551,
+      "step": 5938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.937479126921311e-05,
+      "loss": 2.2459,
+      "step": 5939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9374582579394063e-05,
+      "loss": 2.0013,
+      "step": 5940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9374373855875614e-05,
+      "loss": 2.344,
+      "step": 5941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.93741650986585e-05,
+      "loss": 1.8188,
+      "step": 5942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9373956307743486e-05,
+      "loss": 1.7766,
+      "step": 5943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9373747483131312e-05,
+      "loss": 1.9224,
+      "step": 5944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9373538624822738e-05,
+      "loss": 1.9645,
+      "step": 5945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9373329732818502e-05,
+      "loss": 1.9667,
+      "step": 5946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9373120807119366e-05,
+      "loss": 2.0632,
+      "step": 5947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9372911847726077e-05,
+      "loss": 1.9875,
+      "step": 5948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.937270285463939e-05,
+      "loss": 2.1818,
+      "step": 5949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9372493827860047e-05,
+      "loss": 1.6242,
+      "step": 5950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9372284767388808e-05,
+      "loss": 1.7288,
+      "step": 5951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9372075673226423e-05,
+      "loss": 2.0623,
+      "step": 5952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9371866545373642e-05,
+      "loss": 1.8908,
+      "step": 5953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9371657383831213e-05,
+      "loss": 1.9399,
+      "step": 5954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.93714481885999e-05,
+      "loss": 1.8429,
+      "step": 5955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9371238959680437e-05,
+      "loss": 1.6518,
+      "step": 5956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9371029697073595e-05,
+      "loss": 1.8412,
+      "step": 5957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.937082040078011e-05,
+      "loss": 1.8409,
+      "step": 5958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9370611070800747e-05,
+      "loss": 1.7238,
+      "step": 5959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9370401707136253e-05,
+      "loss": 2.1261,
+      "step": 5960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.937019230978738e-05,
+      "loss": 1.5881,
+      "step": 5961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.936998287875488e-05,
+      "loss": 1.9343,
+      "step": 5962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.936977341403951e-05,
+      "loss": 1.8203,
+      "step": 5963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.936956391564202e-05,
+      "loss": 2.1727,
+      "step": 5964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.936935438356316e-05,
+      "loss": 1.956,
+      "step": 5965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9369144817803695e-05,
+      "loss": 1.6534,
+      "step": 5966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.936893521836436e-05,
+      "loss": 1.6521,
+      "step": 5967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9368725585245927e-05,
+      "loss": 1.6568,
+      "step": 5968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9368515918449135e-05,
+      "loss": 2.1988,
+      "step": 5969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9368306217974746e-05,
+      "loss": 1.8733,
+      "step": 5970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9368096483823513e-05,
+      "loss": 1.5952,
+      "step": 5971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9367886715996188e-05,
+      "loss": 1.6084,
+      "step": 5972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9367676914493526e-05,
+      "loss": 2.0976,
+      "step": 5973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.936746707931628e-05,
+      "loss": 1.9678,
+      "step": 5974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9367257210465206e-05,
+      "loss": 1.8262,
+      "step": 5975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9367047307941055e-05,
+      "loss": 1.7885,
+      "step": 5976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.936683737174459e-05,
+      "loss": 1.9935,
+      "step": 5977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9366627401876553e-05,
+      "loss": 1.8579,
+      "step": 5978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9366417398337712e-05,
+      "loss": 1.7203,
+      "step": 5979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.936620736112881e-05,
+      "loss": 2.0811,
+      "step": 5980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.936599729025061e-05,
+      "loss": 2.0757,
+      "step": 5981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9365787185703867e-05,
+      "loss": 2.0046,
+      "step": 5982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.936557704748933e-05,
+      "loss": 1.777,
+      "step": 5983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.936536687560776e-05,
+      "loss": 2.1074,
+      "step": 5984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.936515667005991e-05,
+      "loss": 2.089,
+      "step": 5985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9364946430846538e-05,
+      "loss": 1.9904,
+      "step": 5986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.93647361579684e-05,
+      "loss": 1.7094,
+      "step": 5987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.936452585142625e-05,
+      "loss": 1.8081,
+      "step": 5988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9364315511220838e-05,
+      "loss": 1.7449,
+      "step": 5989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.936410513735293e-05,
+      "loss": 1.8357,
+      "step": 5990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9363894729823282e-05,
+      "loss": 1.8131,
+      "step": 5991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9363684288632645e-05,
+      "loss": 2.0402,
+      "step": 5992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9363473813781776e-05,
+      "loss": 1.8451,
+      "step": 5993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9363263305271433e-05,
+      "loss": 2.0852,
+      "step": 5994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9363052763102373e-05,
+      "loss": 2.0387,
+      "step": 5995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9362842187275354e-05,
+      "loss": 2.0632,
+      "step": 5996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9362631577791133e-05,
+      "loss": 2.1616,
+      "step": 5997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.936242093465046e-05,
+      "loss": 2.0526,
+      "step": 5998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.93622102578541e-05,
+      "loss": 1.9374,
+      "step": 5999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9361999547402814e-05,
+      "loss": 1.829,
+      "step": 6000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.936178880329735e-05,
+      "loss": 1.9274,
+      "step": 6001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9361578025538464e-05,
+      "loss": 1.6798,
+      "step": 6002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9361367214126924e-05,
+      "loss": 2.1256,
+      "step": 6003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9361156369063483e-05,
+      "loss": 1.8709,
+      "step": 6004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.93609454903489e-05,
+      "loss": 1.8752,
+      "step": 6005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9360734577983925e-05,
+      "loss": 1.9803,
+      "step": 6006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.936052363196933e-05,
+      "loss": 1.5074,
+      "step": 6007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9360312652305864e-05,
+      "loss": 1.7397,
+      "step": 6008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.936010163899429e-05,
+      "loss": 1.9631,
+      "step": 6009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.935989059203536e-05,
+      "loss": 1.739,
+      "step": 6010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.935967951142984e-05,
+      "loss": 2.0956,
+      "step": 6011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.935946839717848e-05,
+      "loss": 1.8139,
+      "step": 6012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9359257249282052e-05,
+      "loss": 1.7618,
+      "step": 6013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9359046067741302e-05,
+      "loss": 2.3475,
+      "step": 6014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9358834852557e-05,
+      "loss": 2.0421,
+      "step": 6015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9358623603729897e-05,
+      "loss": 1.7801,
+      "step": 6016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9358412321260752e-05,
+      "loss": 1.8239,
+      "step": 6017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9358201005150333e-05,
+      "loss": 1.7878,
+      "step": 6018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9357989655399393e-05,
+      "loss": 1.9929,
+      "step": 6019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9357778272008693e-05,
+      "loss": 2.131,
+      "step": 6020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9357566854978993e-05,
+      "loss": 1.7989,
+      "step": 6021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9357355404311054e-05,
+      "loss": 1.6907,
+      "step": 6022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9357143920005638e-05,
+      "loss": 2.0174,
+      "step": 6023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9356932402063496e-05,
+      "loss": 2.0693,
+      "step": 6024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9356720850485398e-05,
+      "loss": 1.9865,
+      "step": 6025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.93565092652721e-05,
+      "loss": 1.8867,
+      "step": 6026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9356297646424366e-05,
+      "loss": 2.1264,
+      "step": 6027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9356085993942953e-05,
+      "loss": 2.0401,
+      "step": 6028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9355874307828627e-05,
+      "loss": 1.9691,
+      "step": 6029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9355662588082142e-05,
+      "loss": 1.9504,
+      "step": 6030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9355450834704265e-05,
+      "loss": 1.9432,
+      "step": 6031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.935523904769575e-05,
+      "loss": 2.2021,
+      "step": 6032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.935502722705737e-05,
+      "loss": 1.8328,
+      "step": 6033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9354815372789874e-05,
+      "loss": 1.8383,
+      "step": 6034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9354603484894034e-05,
+      "loss": 1.9126,
+      "step": 6035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9354391563370603e-05,
+      "loss": 2.0221,
+      "step": 6036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9354179608220347e-05,
+      "loss": 2.0324,
+      "step": 6037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9353967619444028e-05,
+      "loss": 2.0186,
+      "step": 6038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9353755597042406e-05,
+      "loss": 2.3193,
+      "step": 6039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.935354354101625e-05,
+      "loss": 1.8048,
+      "step": 6040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9353331451366312e-05,
+      "loss": 2.1446,
+      "step": 6041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9353119328093362e-05,
+      "loss": 2.0824,
+      "step": 6042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9352907171198157e-05,
+      "loss": 1.661,
+      "step": 6043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9352694980681464e-05,
+      "loss": 1.6334,
+      "step": 6044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9352482756544044e-05,
+      "loss": 1.8143,
+      "step": 6045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.935227049878666e-05,
+      "loss": 1.8565,
+      "step": 6046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9352058207410077e-05,
+      "loss": 1.5292,
+      "step": 6047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9351845882415058e-05,
+      "loss": 2.0071,
+      "step": 6048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.935163352380236e-05,
+      "loss": 1.8766,
+      "step": 6049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9351421131572755e-05,
+      "loss": 1.8053,
+      "step": 6050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9351208705727e-05,
+      "loss": 1.934,
+      "step": 6051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.935099624626586e-05,
+      "loss": 1.625,
+      "step": 6052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9350783753190104e-05,
+      "loss": 2.1696,
+      "step": 6053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.935057122650049e-05,
+      "loss": 1.8066,
+      "step": 6054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9350358666197785e-05,
+      "loss": 2.3901,
+      "step": 6055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.935014607228275e-05,
+      "loss": 1.9546,
+      "step": 6056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9349933444756153e-05,
+      "loss": 1.9686,
+      "step": 6057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9349720783618755e-05,
+      "loss": 1.7126,
+      "step": 6058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9349508088871324e-05,
+      "loss": 1.9882,
+      "step": 6059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9349295360514618e-05,
+      "loss": 2.0443,
+      "step": 6060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.934908259854941e-05,
+      "loss": 2.0139,
+      "step": 6061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.934886980297646e-05,
+      "loss": 1.8877,
+      "step": 6062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9348656973796533e-05,
+      "loss": 1.9201,
+      "step": 6063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9348444111010396e-05,
+      "loss": 2.2813,
+      "step": 6064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9348231214618813e-05,
+      "loss": 1.9574,
+      "step": 6065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9348018284622553e-05,
+      "loss": 1.7391,
+      "step": 6066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.934780532102237e-05,
+      "loss": 1.9154,
+      "step": 6067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9347592323819048e-05,
+      "loss": 2.032,
+      "step": 6068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9347379293013335e-05,
+      "loss": 2.0718,
+      "step": 6069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9347166228606004e-05,
+      "loss": 1.7998,
+      "step": 6070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9346953130597824e-05,
+      "loss": 1.8018,
+      "step": 6071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9346739998989556e-05,
+      "loss": 1.8367,
+      "step": 6072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.934652683378197e-05,
+      "loss": 1.9686,
+      "step": 6073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.934631363497583e-05,
+      "loss": 1.7433,
+      "step": 6074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9346100402571898e-05,
+      "loss": 1.5808,
+      "step": 6075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9345887136570952e-05,
+      "loss": 2.0701,
+      "step": 6076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.934567383697375e-05,
+      "loss": 1.9544,
+      "step": 6077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.934546050378106e-05,
+      "loss": 1.9471,
+      "step": 6078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.934524713699365e-05,
+      "loss": 1.8052,
+      "step": 6079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9345033736612288e-05,
+      "loss": 2.0755,
+      "step": 6080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9344820302637736e-05,
+      "loss": 2.0702,
+      "step": 6081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9344606835070766e-05,
+      "loss": 1.9365,
+      "step": 6082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9344393333912146e-05,
+      "loss": 2.0591,
+      "step": 6083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9344179799162643e-05,
+      "loss": 1.8143,
+      "step": 6084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.934396623082302e-05,
+      "loss": 1.8411,
+      "step": 6085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9343752628894053e-05,
+      "loss": 2.1096,
+      "step": 6086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.93435389933765e-05,
+      "loss": 1.9219,
+      "step": 6087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9343325324271138e-05,
+      "loss": 2.0233,
+      "step": 6088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9343111621578728e-05,
+      "loss": 1.8419,
+      "step": 6089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9342897885300042e-05,
+      "loss": 1.6075,
+      "step": 6090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.934268411543585e-05,
+      "loss": 1.9718,
+      "step": 6091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9342470311986915e-05,
+      "loss": 2.398,
+      "step": 6092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.934225647495401e-05,
+      "loss": 2.127,
+      "step": 6093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.93420426043379e-05,
+      "loss": 1.6105,
+      "step": 6094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.934182870013936e-05,
+      "loss": 1.9638,
+      "step": 6095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9341614762359153e-05,
+      "loss": 2.018,
+      "step": 6096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.934140079099805e-05,
+      "loss": 1.7926,
+      "step": 6097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9341186786056823e-05,
+      "loss": 1.578,
+      "step": 6098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9340972747536236e-05,
+      "loss": 1.6301,
+      "step": 6099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.934075867543706e-05,
+      "loss": 1.9363,
+      "step": 6100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9340544569760065e-05,
+      "loss": 2.2366,
+      "step": 6101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9340330430506027e-05,
+      "loss": 1.686,
+      "step": 6102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.934011625767571e-05,
+      "loss": 1.6773,
+      "step": 6103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9339902051269877e-05,
+      "loss": 1.8889,
+      "step": 6104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9339687811289307e-05,
+      "loss": 1.8718,
+      "step": 6105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9339473537734773e-05,
+      "loss": 2.1755,
+      "step": 6106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9339259230607035e-05,
+      "loss": 1.5395,
+      "step": 6107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.933904488990687e-05,
+      "loss": 1.5314,
+      "step": 6108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9338830515635048e-05,
+      "loss": 1.967,
+      "step": 6109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9338616107792337e-05,
+      "loss": 2.0575,
+      "step": 6110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9338401666379512e-05,
+      "loss": 2.5134,
+      "step": 6111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9338187191397342e-05,
+      "loss": 1.79,
+      "step": 6112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.93379726828466e-05,
+      "loss": 1.6215,
+      "step": 6113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9337758140728048e-05,
+      "loss": 1.864,
+      "step": 6114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9337543565042465e-05,
+      "loss": 2.0117,
+      "step": 6115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.933732895579062e-05,
+      "loss": 1.9832,
+      "step": 6116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9337114312973294e-05,
+      "loss": 1.7915,
+      "step": 6117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9336899636591242e-05,
+      "loss": 1.745,
+      "step": 6118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.933668492664525e-05,
+      "loss": 2.0322,
+      "step": 6119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9336470183136077e-05,
+      "loss": 1.833,
+      "step": 6120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9336255406064505e-05,
+      "loss": 1.5688,
+      "step": 6121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9336040595431302e-05,
+      "loss": 1.6631,
+      "step": 6122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.933582575123724e-05,
+      "loss": 1.76,
+      "step": 6123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.93356108734831e-05,
+      "loss": 1.9908,
+      "step": 6124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9335395962169635e-05,
+      "loss": 1.8413,
+      "step": 6125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9335181017297638e-05,
+      "loss": 1.5921,
+      "step": 6126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.933496603886787e-05,
+      "loss": 1.8718,
+      "step": 6127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9334751026881108e-05,
+      "loss": 1.8008,
+      "step": 6128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.933453598133812e-05,
+      "loss": 1.6721,
+      "step": 6129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.933432090223969e-05,
+      "loss": 2.2122,
+      "step": 6130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9334105789586575e-05,
+      "loss": 1.938,
+      "step": 6131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.933389064337956e-05,
+      "loss": 2.099,
+      "step": 6132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9333675463619417e-05,
+      "loss": 1.831,
+      "step": 6133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9333460250306917e-05,
+      "loss": 1.8922,
+      "step": 6134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9333245003442836e-05,
+      "loss": 1.7692,
+      "step": 6135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9333029723027946e-05,
+      "loss": 1.8081,
+      "step": 6136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9332814409063023e-05,
+      "loss": 1.9744,
+      "step": 6137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9332599061548835e-05,
+      "loss": 1.87,
+      "step": 6138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9332383680486163e-05,
+      "loss": 2.1607,
+      "step": 6139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.933216826587578e-05,
+      "loss": 2.1959,
+      "step": 6140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9331952817718454e-05,
+      "loss": 1.9336,
+      "step": 6141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.933173733601497e-05,
+      "loss": 1.7491,
+      "step": 6142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9331521820766097e-05,
+      "loss": 2.098,
+      "step": 6143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.933130627197261e-05,
+      "loss": 1.9409,
+      "step": 6144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.933109068963528e-05,
+      "loss": 1.8696,
+      "step": 6145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.933087507375489e-05,
+      "loss": 1.9078,
+      "step": 6146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.933065942433221e-05,
+      "loss": 1.9228,
+      "step": 6147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.933044374136801e-05,
+      "loss": 1.8746,
+      "step": 6148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9330228024863077e-05,
+      "loss": 2.096,
+      "step": 6149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.933001227481818e-05,
+      "loss": 1.6551,
+      "step": 6150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9329796491234096e-05,
+      "loss": 1.8068,
+      "step": 6151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.93295806741116e-05,
+      "loss": 1.6885,
+      "step": 6152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9329364823451465e-05,
+      "loss": 1.9859,
+      "step": 6153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9329148939254475e-05,
+      "loss": 1.8641,
+      "step": 6154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9328933021521394e-05,
+      "loss": 3.1731,
+      "step": 6155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.932871707025301e-05,
+      "loss": 1.71,
+      "step": 6156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9328501085450092e-05,
+      "loss": 1.5703,
+      "step": 6157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.932828506711342e-05,
+      "loss": 2.0776,
+      "step": 6158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9328069015243768e-05,
+      "loss": 1.8801,
+      "step": 6159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9327852929841918e-05,
+      "loss": 1.9433,
+      "step": 6160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9327636810908638e-05,
+      "loss": 1.6366,
+      "step": 6161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9327420658444713e-05,
+      "loss": 1.7196,
+      "step": 6162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9327204472450917e-05,
+      "loss": 1.9813,
+      "step": 6163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.932698825292802e-05,
+      "loss": 1.8132,
+      "step": 6164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9326771999876814e-05,
+      "loss": 1.8771,
+      "step": 6165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9326555713298062e-05,
+      "loss": 1.8851,
+      "step": 6166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5,
+      "learning_rate": 1.932633939319255e-05,
+      "loss": 2.2393,
+      "step": 6167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9326123039561054e-05,
+      "loss": 1.5964,
+      "step": 6168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9325906652404353e-05,
+      "loss": 2.04,
+      "step": 6169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9325690231723223e-05,
+      "loss": 1.9176,
+      "step": 6170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.932547377751844e-05,
+      "loss": 1.6358,
+      "step": 6171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9325257289790786e-05,
+      "loss": 1.6833,
+      "step": 6172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9325040768541035e-05,
+      "loss": 2.0759,
+      "step": 6173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9324824213769968e-05,
+      "loss": 2.199,
+      "step": 6174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9324607625478365e-05,
+      "loss": 1.7797,
+      "step": 6175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9324391003667005e-05,
+      "loss": 1.6511,
+      "step": 6176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.932417434833666e-05,
+      "loss": 2.2501,
+      "step": 6177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9323957659488112e-05,
+      "loss": 1.7635,
+      "step": 6178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9323740937122147e-05,
+      "loss": 2.2567,
+      "step": 6179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9323524181239536e-05,
+      "loss": 1.7289,
+      "step": 6180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.932330739184106e-05,
+      "loss": 1.8715,
+      "step": 6181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9323090568927498e-05,
+      "loss": 2.1385,
+      "step": 6182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.932287371249963e-05,
+      "loss": 1.7873,
+      "step": 6183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9322656822558235e-05,
+      "loss": 2.0148,
+      "step": 6184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.93224398991041e-05,
+      "loss": 1.9754,
+      "step": 6185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9322222942137993e-05,
+      "loss": 1.8119,
+      "step": 6186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.93220059516607e-05,
+      "loss": 1.986,
+      "step": 6187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9321788927672997e-05,
+      "loss": 1.7335,
+      "step": 6188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9321571870175672e-05,
+      "loss": 1.9141,
+      "step": 6189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.93213547791695e-05,
+      "loss": 1.8403,
+      "step": 6190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.932113765465526e-05,
+      "loss": 2.058,
+      "step": 6191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9320920496633735e-05,
+      "loss": 2.0755,
+      "step": 6192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9320703305105707e-05,
+      "loss": 1.9943,
+      "step": 6193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9320486080071953e-05,
+      "loss": 2.1433,
+      "step": 6194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9320268821533255e-05,
+      "loss": 1.8074,
+      "step": 6195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9320051529490396e-05,
+      "loss": 1.9395,
+      "step": 6196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9319834203944155e-05,
+      "loss": 2.2522,
+      "step": 6197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9319616844895318e-05,
+      "loss": 1.8601,
+      "step": 6198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9319399452344658e-05,
+      "loss": 2.033,
+      "step": 6199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9319182026292964e-05,
+      "loss": 2.3614,
+      "step": 6200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.931896456674101e-05,
+      "loss": 1.9684,
+      "step": 6201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9318747073689587e-05,
+      "loss": 1.7884,
+      "step": 6202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.931852954713947e-05,
+      "loss": 2.0196,
+      "step": 6203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9318311987091442e-05,
+      "loss": 1.6385,
+      "step": 6204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.931809439354629e-05,
+      "loss": 1.8547,
+      "step": 6205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9317876766504788e-05,
+      "loss": 1.8745,
+      "step": 6206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.931765910596773e-05,
+      "loss": 1.9266,
+      "step": 6207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.931744141193588e-05,
+      "loss": 1.8073,
+      "step": 6208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.931722368441004e-05,
+      "loss": 1.8339,
+      "step": 6209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.931700592339098e-05,
+      "loss": 1.9717,
+      "step": 6210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.931678812887949e-05,
+      "loss": 1.7977,
+      "step": 6211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9316570300876347e-05,
+      "loss": 1.9117,
+      "step": 6212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9316352439382338e-05,
+      "loss": 1.9871,
+      "step": 6213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9316134544398246e-05,
+      "loss": 2.0058,
+      "step": 6214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9315916615924856e-05,
+      "loss": 1.7301,
+      "step": 6215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9315698653962944e-05,
+      "loss": 1.905,
+      "step": 6216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.93154806585133e-05,
+      "loss": 2.0222,
+      "step": 6217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.931526262957671e-05,
+      "loss": 1.8967,
+      "step": 6218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.931504456715395e-05,
+      "loss": 1.9935,
+      "step": 6219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.931482647124581e-05,
+      "loss": 1.9509,
+      "step": 6220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.931460834185307e-05,
+      "loss": 1.9391,
+      "step": 6221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9314390178976515e-05,
+      "loss": 2.026,
+      "step": 6222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.931417198261693e-05,
+      "loss": 1.9796,
+      "step": 6223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9313953752775104e-05,
+      "loss": 1.6765,
+      "step": 6224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9313735489451812e-05,
+      "loss": 2.1128,
+      "step": 6225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9313517192647846e-05,
+      "loss": 1.9114,
+      "step": 6226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9313298862363984e-05,
+      "loss": 1.9378,
+      "step": 6227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.931308049860102e-05,
+      "loss": 1.8819,
+      "step": 6228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9312862101359732e-05,
+      "loss": 2.0165,
+      "step": 6229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.931264367064091e-05,
+      "loss": 1.8858,
+      "step": 6230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9312425206445334e-05,
+      "loss": 2.0522,
+      "step": 6231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.931220670877379e-05,
+      "loss": 1.9732,
+      "step": 6232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.931198817762707e-05,
+      "loss": 1.99,
+      "step": 6233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9311769613005952e-05,
+      "loss": 1.9507,
+      "step": 6234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.931155101491122e-05,
+      "loss": 1.6897,
+      "step": 6235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.931133238334367e-05,
+      "loss": 2.0478,
+      "step": 6236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0,
+      "learning_rate": 1.931111371830408e-05,
+      "loss": 2.5812,
+      "step": 6237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.931089501979324e-05,
+      "loss": 1.9327,
+      "step": 6238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9310676287811935e-05,
+      "loss": 2.0439,
+      "step": 6239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9310457522360946e-05,
+      "loss": 1.7465,
+      "step": 6240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.931023872344107e-05,
+      "loss": 2.1114,
+      "step": 6241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9310019891053082e-05,
+      "loss": 1.9582,
+      "step": 6242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9309801025197775e-05,
+      "loss": 1.8622,
+      "step": 6243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.930958212587594e-05,
+      "loss": 1.5794,
+      "step": 6244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9309363193088354e-05,
+      "loss": 1.6209,
+      "step": 6245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.930914422683581e-05,
+      "loss": 1.9503,
+      "step": 6246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9308925227119094e-05,
+      "loss": 1.8106,
+      "step": 6247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9308706193938995e-05,
+      "loss": 1.7549,
+      "step": 6248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9308487127296297e-05,
+      "loss": 2.3355,
+      "step": 6249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.930826802719179e-05,
+      "loss": 1.9595,
+      "step": 6250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.930804889362626e-05,
+      "loss": 1.7789,
+      "step": 6251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9307829726600496e-05,
+      "loss": 2.0859,
+      "step": 6252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9307610526115286e-05,
+      "loss": 2.01,
+      "step": 6253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9307391292171418e-05,
+      "loss": 1.9585,
+      "step": 6254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9307172024769677e-05,
+      "loss": 1.9277,
+      "step": 6255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9306952723910856e-05,
+      "loss": 1.9959,
+      "step": 6256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.930673338959574e-05,
+      "loss": 1.9195,
+      "step": 6257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.930651402182512e-05,
+      "loss": 1.7735,
+      "step": 6258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.930629462059978e-05,
+      "loss": 1.9213,
+      "step": 6259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9306075185920513e-05,
+      "loss": 1.8477,
+      "step": 6260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9305855717788104e-05,
+      "loss": 1.8437,
+      "step": 6261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.930563621620335e-05,
+      "loss": 1.964,
+      "step": 6262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.930541668116703e-05,
+      "loss": 1.9396,
+      "step": 6263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9305197112679937e-05,
+      "loss": 1.7318,
+      "step": 6264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9304977510742865e-05,
+      "loss": 2.2352,
+      "step": 6265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.93047578753566e-05,
+      "loss": 1.8334,
+      "step": 6266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.930453820652193e-05,
+      "loss": 1.853,
+      "step": 6267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9304318504239642e-05,
+      "loss": 1.6229,
+      "step": 6268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.930409876851053e-05,
+      "loss": 1.7456,
+      "step": 6269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9303878999335387e-05,
+      "loss": 2.7466,
+      "step": 6270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9303659196714997e-05,
+      "loss": 1.9762,
+      "step": 6271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.930343936065015e-05,
+      "loss": 2.0105,
+      "step": 6272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9303219491141643e-05,
+      "loss": 1.6857,
+      "step": 6273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.930299958819026e-05,
+      "loss": 1.7225,
+      "step": 6274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9302779651796796e-05,
+      "loss": 1.8698,
+      "step": 6275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9302559681962037e-05,
+      "loss": 2.2839,
+      "step": 6276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9302339678686775e-05,
+      "loss": 1.8723,
+      "step": 6277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9302119641971805e-05,
+      "loss": 1.7214,
+      "step": 6278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9301899571817912e-05,
+      "loss": 2.0932,
+      "step": 6279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.930167946822589e-05,
+      "loss": 1.6589,
+      "step": 6280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.930145933119653e-05,
+      "loss": 1.6418,
+      "step": 6281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9301239160730625e-05,
+      "loss": 1.8518,
+      "step": 6282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9301018956828966e-05,
+      "loss": 2.0047,
+      "step": 6283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9300798719492338e-05,
+      "loss": 1.9933,
+      "step": 6284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9300578448721544e-05,
+      "loss": 1.7324,
+      "step": 6285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9300358144517367e-05,
+      "loss": 1.8545,
+      "step": 6286
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9300137806880604e-05,
+      "loss": 2.1739,
+      "step": 6287
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9299917435812042e-05,
+      "loss": 1.8622,
+      "step": 6288
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.929969703131248e-05,
+      "loss": 1.681,
+      "step": 6289
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9299476593382702e-05,
+      "loss": 1.8164,
+      "step": 6290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9299256122023505e-05,
+      "loss": 1.8386,
+      "step": 6291
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9299035617235685e-05,
+      "loss": 1.8446,
+      "step": 6292
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.929881507902003e-05,
+      "loss": 1.7454,
+      "step": 6293
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9298594507377335e-05,
+      "loss": 1.6081,
+      "step": 6294
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.929837390230839e-05,
+      "loss": 1.9567,
+      "step": 6295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.929815326381399e-05,
+      "loss": 1.8106,
+      "step": 6296
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.929793259189493e-05,
+      "loss": 1.705,
+      "step": 6297
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9297711886551998e-05,
+      "loss": 2.3759,
+      "step": 6298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9297491147785994e-05,
+      "loss": 2.0118,
+      "step": 6299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9297270375597703e-05,
+      "loss": 2.0624,
+      "step": 6300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9297049569987927e-05,
+      "loss": 1.899,
+      "step": 6301
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9296828730957458e-05,
+      "loss": 2.1085,
+      "step": 6302
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.929660785850709e-05,
+      "loss": 2.2508,
+      "step": 6303
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.929638695263761e-05,
+      "loss": 1.6968,
+      "step": 6304
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.929616601334982e-05,
+      "loss": 1.9279,
+      "step": 6305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.929594504064451e-05,
+      "loss": 2.0453,
+      "step": 6306
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9295724034522476e-05,
+      "loss": 1.9844,
+      "step": 6307
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9295502994984515e-05,
+      "loss": 1.858,
+      "step": 6308
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9295281922031416e-05,
+      "loss": 2.2801,
+      "step": 6309
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9295060815663978e-05,
+      "loss": 1.9337,
+      "step": 6310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9294839675883e-05,
+      "loss": 1.8914,
+      "step": 6311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9294618502689266e-05,
+      "loss": 1.9506,
+      "step": 6312
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9294397296083573e-05,
+      "loss": 1.6674,
+      "step": 6313
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9294176056066724e-05,
+      "loss": 1.9134,
+      "step": 6314
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9293954782639513e-05,
+      "loss": 2.0415,
+      "step": 6315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9293733475802728e-05,
+      "loss": 1.6331,
+      "step": 6316
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9293512135557175e-05,
+      "loss": 1.8339,
+      "step": 6317
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9293290761903635e-05,
+      "loss": 1.7865,
+      "step": 6318
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.929306935484292e-05,
+      "loss": 1.8544,
+      "step": 6319
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9292847914375815e-05,
+      "loss": 1.6708,
+      "step": 6320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.929262644050312e-05,
+      "loss": 1.6046,
+      "step": 6321
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.929240493322563e-05,
+      "loss": 2.1311,
+      "step": 6322
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9292183392544143e-05,
+      "loss": 1.8832,
+      "step": 6323
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9291961818459454e-05,
+      "loss": 2.0961,
+      "step": 6324
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.929174021097236e-05,
+      "loss": 1.9388,
+      "step": 6325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9291518570083652e-05,
+      "loss": 1.7985,
+      "step": 6326
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.929129689579414e-05,
+      "loss": 1.9161,
+      "step": 6327
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9291075188104605e-05,
+      "loss": 2.3963,
+      "step": 6328
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9290853447015858e-05,
+      "loss": 1.6436,
+      "step": 6329
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.929063167252869e-05,
+      "loss": 2.0907,
+      "step": 6330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9290409864643896e-05,
+      "loss": 1.925,
+      "step": 6331
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.929018802336228e-05,
+      "loss": 1.9124,
+      "step": 6332
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.928996614868463e-05,
+      "loss": 1.5858,
+      "step": 6333
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.928974424061175e-05,
+      "loss": 1.9722,
+      "step": 6334
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.928952229914444e-05,
+      "loss": 1.8871,
+      "step": 6335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.928930032428349e-05,
+      "loss": 1.9954,
+      "step": 6336
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9289078316029705e-05,
+      "loss": 1.4969,
+      "step": 6337
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9288856274383876e-05,
+      "loss": 2.0611,
+      "step": 6338
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.928863419934681e-05,
+      "loss": 1.8362,
+      "step": 6339
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9288412090919296e-05,
+      "loss": 1.6381,
+      "step": 6340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.928818994910214e-05,
+      "loss": 2.1202,
+      "step": 6341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.928796777389614e-05,
+      "loss": 1.8783,
+      "step": 6342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.928774556530209e-05,
+      "loss": 1.6908,
+      "step": 6343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.928752332332079e-05,
+      "loss": 1.6212,
+      "step": 6344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9287301047953043e-05,
+      "loss": 2.1837,
+      "step": 6345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9287078739199643e-05,
+      "loss": 1.8488,
+      "step": 6346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.928685639706139e-05,
+      "loss": 1.6087,
+      "step": 6347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9286634021539088e-05,
+      "loss": 1.7173,
+      "step": 6348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9286411612633532e-05,
+      "loss": 1.5782,
+      "step": 6349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.928618917034552e-05,
+      "loss": 1.8762,
+      "step": 6350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5,
+      "learning_rate": 1.928596669467586e-05,
+      "loss": 2.0272,
+      "step": 6351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.928574418562534e-05,
+      "loss": 2.0248,
+      "step": 6352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9285521643194768e-05,
+      "loss": 1.734,
+      "step": 6353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.928529906738494e-05,
+      "loss": 1.8074,
+      "step": 6354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9285076458196656e-05,
+      "loss": 1.7947,
+      "step": 6355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9284853815630722e-05,
+      "loss": 2.1872,
+      "step": 6356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9284631139687933e-05,
+      "loss": 1.8288,
+      "step": 6357
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9284408430369092e-05,
+      "loss": 1.792,
+      "step": 6358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9284185687674997e-05,
+      "loss": 2.0863,
+      "step": 6359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.928396291160645e-05,
+      "loss": 1.9338,
+      "step": 6360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.928374010216425e-05,
+      "loss": 1.9739,
+      "step": 6361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9283517259349206e-05,
+      "loss": 2.2315,
+      "step": 6362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.928329438316211e-05,
+      "loss": 1.9846,
+      "step": 6363
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9283071473603764e-05,
+      "loss": 1.7662,
+      "step": 6364
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.928284853067497e-05,
+      "loss": 1.0669,
+      "step": 6365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9282625554376534e-05,
+      "loss": 1.9544,
+      "step": 6366
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9282402544709256e-05,
+      "loss": 1.8501,
+      "step": 6367
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9282179501673934e-05,
+      "loss": 1.948,
+      "step": 6368
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9281956425271372e-05,
+      "loss": 1.7821,
+      "step": 6369
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.928173331550237e-05,
+      "loss": 2.1973,
+      "step": 6370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.928151017236773e-05,
+      "loss": 1.7637,
+      "step": 6371
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9281286995868264e-05,
+      "loss": 2.1416,
+      "step": 6372
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.928106378600476e-05,
+      "loss": 1.9251,
+      "step": 6373
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.928084054277803e-05,
+      "loss": 1.4772,
+      "step": 6374
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.928061726618887e-05,
+      "loss": 1.8056,
+      "step": 6375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9280393956238086e-05,
+      "loss": 1.6978,
+      "step": 6376
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.928017061292648e-05,
+      "loss": 2.0249,
+      "step": 6377
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.927994723625486e-05,
+      "loss": 2.0779,
+      "step": 6378
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9279723826224018e-05,
+      "loss": 1.6415,
+      "step": 6379
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.927950038283477e-05,
+      "loss": 1.7258,
+      "step": 6380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9279276906087906e-05,
+      "loss": 2.0909,
+      "step": 6381
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.927905339598424e-05,
+      "loss": 1.8161,
+      "step": 6382
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9278829852524572e-05,
+      "loss": 2.0033,
+      "step": 6383
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9278606275709704e-05,
+      "loss": 1.9148,
+      "step": 6384
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.927838266554044e-05,
+      "loss": 1.7572,
+      "step": 6385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9278159022017587e-05,
+      "loss": 1.8501,
+      "step": 6386
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.927793534514194e-05,
+      "loss": 2.0234,
+      "step": 6387
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.927771163491432e-05,
+      "loss": 1.9886,
+      "step": 6388
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9277487891335512e-05,
+      "loss": 1.6296,
+      "step": 6389
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9277264114406332e-05,
+      "loss": 1.7667,
+      "step": 6390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9277040304127584e-05,
+      "loss": 2.0008,
+      "step": 6391
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.927681646050007e-05,
+      "loss": 1.9181,
+      "step": 6392
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9276592583524588e-05,
+      "loss": 1.5803,
+      "step": 6393
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9276368673201955e-05,
+      "loss": 1.9362,
+      "step": 6394
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9276144729532967e-05,
+      "loss": 1.8449,
+      "step": 6395
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9275920752518437e-05,
+      "loss": 2.0962,
+      "step": 6396
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9275696742159164e-05,
+      "loss": 1.9608,
+      "step": 6397
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9275472698455954e-05,
+      "loss": 2.0172,
+      "step": 6398
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9275248621409614e-05,
+      "loss": 2.316,
+      "step": 6399
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.927502451102095e-05,
+      "loss": 1.585,
+      "step": 6400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9274800367290763e-05,
+      "loss": 1.9826,
+      "step": 6401
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9274576190219864e-05,
+      "loss": 2.0195,
+      "step": 6402
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9274351979809057e-05,
+      "loss": 1.9894,
+      "step": 6403
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.927412773605915e-05,
+      "loss": 2.0118,
+      "step": 6404
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9273903458970942e-05,
+      "loss": 2.4985,
+      "step": 6405
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9273679148545246e-05,
+      "loss": 1.8241,
+      "step": 6406
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.927345480478287e-05,
+      "loss": 1.7822,
+      "step": 6407
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9273230427684613e-05,
+      "loss": 2.1747,
+      "step": 6408
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.927300601725129e-05,
+      "loss": 1.8955,
+      "step": 6409
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.92727815734837e-05,
+      "loss": 2.1709,
+      "step": 6410
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9272557096382654e-05,
+      "loss": 2.0105,
+      "step": 6411
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9272332585948956e-05,
+      "loss": 2.0926,
+      "step": 6412
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.927210804218342e-05,
+      "loss": 1.805,
+      "step": 6413
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9271883465086843e-05,
+      "loss": 2.2313,
+      "step": 6414
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9271658854660044e-05,
+      "loss": 1.5824,
+      "step": 6415
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9271434210903817e-05,
+      "loss": 1.8327,
+      "step": 6416
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.927120953381898e-05,
+      "loss": 2.0256,
+      "step": 6417
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9270984823406337e-05,
+      "loss": 1.898,
+      "step": 6418
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9270760079666697e-05,
+      "loss": 1.7809,
+      "step": 6419
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9270535302600867e-05,
+      "loss": 1.7649,
+      "step": 6420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9270310492209655e-05,
+      "loss": 1.8603,
+      "step": 6421
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.927008564849387e-05,
+      "loss": 1.842,
+      "step": 6422
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9269860771454315e-05,
+      "loss": 1.7207,
+      "step": 6423
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.9269635861091805e-05,
+      "loss": 2.1812,
+      "step": 6424
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.926941091740715e-05,
+      "loss": 1.8426,
+      "step": 6425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9269185940401153e-05,
+      "loss": 1.9308,
+      "step": 6426
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9268960930074623e-05,
+      "loss": 1.9387,
+      "step": 6427
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9268735886428368e-05,
+      "loss": 1.9313,
+      "step": 6428
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9268510809463205e-05,
+      "loss": 1.7823,
+      "step": 6429
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9268285699179933e-05,
+      "loss": 2.0693,
+      "step": 6430
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9268060555579373e-05,
+      "loss": 2.1629,
+      "step": 6431
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.926783537866232e-05,
+      "loss": 1.7718,
+      "step": 6432
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.926761016842959e-05,
+      "loss": 1.9181,
+      "step": 6433
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9267384924881998e-05,
+      "loss": 1.9698,
+      "step": 6434
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9267159648020346e-05,
+      "loss": 1.8267,
+      "step": 6435
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.926693433784545e-05,
+      "loss": 1.9939,
+      "step": 6436
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9266708994358114e-05,
+      "loss": 2.183,
+      "step": 6437
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9266483617559148e-05,
+      "loss": 1.6005,
+      "step": 6438
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9266258207449367e-05,
+      "loss": 1.7313,
+      "step": 6439
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9266032764029582e-05,
+      "loss": 2.309,
+      "step": 6440
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.92658072873006e-05,
+      "loss": 1.7334,
+      "step": 6441
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9265581777263228e-05,
+      "loss": 1.9887,
+      "step": 6442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9265356233918286e-05,
+      "loss": 1.7498,
+      "step": 6443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9265130657266574e-05,
+      "loss": 1.742,
+      "step": 6444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9264905047308914e-05,
+      "loss": 2.2517,
+      "step": 6445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9264679404046108e-05,
+      "loss": 1.9699,
+      "step": 6446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.926445372747897e-05,
+      "loss": 2.0613,
+      "step": 6447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.926422801760831e-05,
+      "loss": 1.8213,
+      "step": 6448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9264002274434947e-05,
+      "loss": 1.8532,
+      "step": 6449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9263776497959683e-05,
+      "loss": 2.069,
+      "step": 6450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9263550688183336e-05,
+      "loss": 2.3321,
+      "step": 6451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.926332484510671e-05,
+      "loss": 1.9212,
+      "step": 6452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.926309896873063e-05,
+      "loss": 2.0239,
+      "step": 6453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9262873059055894e-05,
+      "loss": 1.7853,
+      "step": 6454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.926264711608332e-05,
+      "loss": 1.8516,
+      "step": 6455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9262421139813723e-05,
+      "loss": 1.6628,
+      "step": 6456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.926219513024791e-05,
+      "loss": 1.8326,
+      "step": 6457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9261969087386697e-05,
+      "loss": 1.9698,
+      "step": 6458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9261743011230894e-05,
+      "loss": 1.5819,
+      "step": 6459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9261516901781317e-05,
+      "loss": 1.775,
+      "step": 6460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5,
+      "learning_rate": 1.9261290759038774e-05,
+      "loss": 2.206,
+      "step": 6461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9261064583004086e-05,
+      "loss": 1.6487,
+      "step": 6462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9260838373678055e-05,
+      "loss": 2.2995,
+      "step": 6463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9260612131061506e-05,
+      "loss": 2.1095,
+      "step": 6464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9260385855155243e-05,
+      "loss": 2.2057,
+      "step": 6465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9260159545960087e-05,
+      "loss": 1.7725,
+      "step": 6466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9259933203476842e-05,
+      "loss": 1.7082,
+      "step": 6467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.925970682770633e-05,
+      "loss": 1.7135,
+      "step": 6468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.925948041864936e-05,
+      "loss": 1.8664,
+      "step": 6469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.925925397630675e-05,
+      "loss": 1.9737,
+      "step": 6470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9259027500679312e-05,
+      "loss": 1.8268,
+      "step": 6471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9258800991767855e-05,
+      "loss": 2.0946,
+      "step": 6472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.92585744495732e-05,
+      "loss": 2.0034,
+      "step": 6473
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9258347874096165e-05,
+      "loss": 1.7945,
+      "step": 6474
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9258121265337552e-05,
+      "loss": 1.5701,
+      "step": 6475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9257894623298187e-05,
+      "loss": 1.8358,
+      "step": 6476
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.925766794797888e-05,
+      "loss": 1.96,
+      "step": 6477
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9257441239380444e-05,
+      "loss": 1.9304,
+      "step": 6478
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9257214497503698e-05,
+      "loss": 1.6467,
+      "step": 6479
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9256987722349453e-05,
+      "loss": 1.8798,
+      "step": 6480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9256760913918527e-05,
+      "loss": 2.5933,
+      "step": 6481
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9256534072211735e-05,
+      "loss": 1.7086,
+      "step": 6482
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9256307197229894e-05,
+      "loss": 1.9616,
+      "step": 6483
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9256080288973817e-05,
+      "loss": 1.8784,
+      "step": 6484
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9255853347444317e-05,
+      "loss": 1.7869,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9255626372642215e-05,
+      "loss": 1.8642,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9255399364568323e-05,
+      "loss": 1.8862,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9255172323223465e-05,
+      "loss": 1.9497,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9254945248608447e-05,
+      "loss": 1.8003,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.925471814072409e-05,
+      "loss": 1.7358,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.925449099957121e-05,
+      "loss": 1.6871,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9254263825150622e-05,
+      "loss": 1.9324,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9254036617463147e-05,
+      "loss": 1.9855,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9253809376509597e-05,
+      "loss": 1.7421,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9253582102290794e-05,
+      "loss": 1.7178,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9253354794807547e-05,
+      "loss": 1.7455,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.925312745406068e-05,
+      "loss": 1.8232,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.925290008005101e-05,
+      "loss": 1.8626,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.925267267277935e-05,
+      "loss": 1.6266,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9252445232246522e-05,
+      "loss": 1.9909,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.925221775845334e-05,
+      "loss": 1.6205,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9251990251400623e-05,
+      "loss": 2.1241,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.925176271108919e-05,
+      "loss": 1.9664,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9251535137519857e-05,
+      "loss": 1.7525,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9251307530693445e-05,
+      "loss": 1.9126,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9251079890610768e-05,
+      "loss": 2.0609,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.925085221727265e-05,
+      "loss": 1.9491,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.92506245106799e-05,
+      "loss": 1.8644,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.925039677083335e-05,
+      "loss": 1.7458,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9250168997733803e-05,
+      "loss": 1.8773,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9249941191382092e-05,
+      "loss": 1.7032,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.924971335177903e-05,
+      "loss": 2.0761,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9249485478925434e-05,
+      "loss": 1.97,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9249257572822125e-05,
+      "loss": 2.0488,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9249029633469925e-05,
+      "loss": 1.923,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9248801660869648e-05,
+      "loss": 1.9703,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9248573655022115e-05,
+      "loss": 1.6834,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.924834561592815e-05,
+      "loss": 1.9151,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.924811754358857e-05,
+      "loss": 2.1217,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9247889438004192e-05,
+      "loss": 2.0347,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.924766129917584e-05,
+      "loss": 1.8372,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9247433127104333e-05,
+      "loss": 2.1233,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.924720492179049e-05,
+      "loss": 2.0403,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9246976683235134e-05,
+      "loss": 2.3942,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.924674841143908e-05,
+      "loss": 2.0251,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9246520106403155e-05,
+      "loss": 2.0083,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9246291768128177e-05,
+      "loss": 2.4563,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9246063396614965e-05,
+      "loss": 1.5751,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9245834991864344e-05,
+      "loss": 1.7227,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9245606553877132e-05,
+      "loss": 1.6248,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9245378082654153e-05,
+      "loss": 1.8111,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9245149578196223e-05,
+      "loss": 2.0821,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9244921040504167e-05,
+      "loss": 1.8341,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9244692469578807e-05,
+      "loss": 2.3995,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9244463865420962e-05,
+      "loss": 2.2732,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9244235228031458e-05,
+      "loss": 1.855,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.9244006557411116e-05,
+      "loss": 1.9434,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9243777853560753e-05,
+      "loss": 1.8878,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9243549116481196e-05,
+      "loss": 1.9361,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9243320346173266e-05,
+      "loss": 1.94,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9243091542637787e-05,
+      "loss": 2.0381,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.924286270587558e-05,
+      "loss": 1.5865,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9242633835887462e-05,
+      "loss": 1.6921,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.9242404932674268e-05,
+      "loss": 3.2541,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.924217599623681e-05,
+      "loss": 2.0824,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9241947026575914e-05,
+      "loss": 2.1213,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9241718023692408e-05,
+      "loss": 1.9441,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.924148898758711e-05,
+      "loss": 2.0898,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9241259918260845e-05,
+      "loss": 2.0029,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9241030815714432e-05,
+      "loss": 1.9543,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9240801679948705e-05,
+      "loss": 1.8141,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9240572510964477e-05,
+      "loss": 1.8792,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9240343308762578e-05,
+      "loss": 1.8176,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9240114073343828e-05,
+      "loss": 1.5459,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9239884804709054e-05,
+      "loss": 2.0119,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.923965550285908e-05,
+      "loss": 1.924,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9239426167794732e-05,
+      "loss": 1.9677,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9239196799516828e-05,
+      "loss": 1.984,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.92389673980262e-05,
+      "loss": 1.8508,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9238737963323663e-05,
+      "loss": 2.2121,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9238508495410056e-05,
+      "loss": 2.0795,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.923827899428619e-05,
+      "loss": 2.1912,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.92380494599529e-05,
+      "loss": 2.2954,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9237819892411002e-05,
+      "loss": 1.9617,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9237590291661328e-05,
+      "loss": 1.9973,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9237360657704706e-05,
+      "loss": 1.1896,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.923713099054195e-05,
+      "loss": 1.8052,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9236901290173896e-05,
+      "loss": 1.8961,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9236671556601368e-05,
+      "loss": 2.0144,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9236441789825185e-05,
+      "loss": 1.7596,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9236211989846186e-05,
+      "loss": 1.4933,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9235982156665182e-05,
+      "loss": 1.9187,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.923575229028301e-05,
+      "loss": 1.8578,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9235522390700492e-05,
+      "loss": 2.2523,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9235292457918456e-05,
+      "loss": 1.9199,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9235062491937726e-05,
+      "loss": 1.8077,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.923483249275913e-05,
+      "loss": 1.8291,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9234602460383497e-05,
+      "loss": 2.0217,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9234372394811652e-05,
+      "loss": 1.884,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.923414229604442e-05,
+      "loss": 2.0609,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.923391216408263e-05,
+      "loss": 2.1724,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9233681998927113e-05,
+      "loss": 1.7762,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.923345180057869e-05,
+      "loss": 2.0923,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9233221569038192e-05,
+      "loss": 1.7092,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9232991304306448e-05,
+      "loss": 1.6171,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9232761006384284e-05,
+      "loss": 1.8662,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9232530675272522e-05,
+      "loss": 2.2536,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9232300310972004e-05,
+      "loss": 1.5271,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9232069913483543e-05,
+      "loss": 1.8101,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9231839482807978e-05,
+      "loss": 2.0792,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.923160901894613e-05,
+      "loss": 1.8095,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9231378521898832e-05,
+      "loss": 1.8413,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.923114799166691e-05,
+      "loss": 1.8853,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.92309174282512e-05,
+      "loss": 1.9261,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.923068683165252e-05,
+      "loss": 1.385,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9230456201871707e-05,
+      "loss": 2.2217,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.923022553890958e-05,
+      "loss": 1.9863,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9229994842766985e-05,
+      "loss": 1.7645,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9229764113444734e-05,
+      "loss": 1.953,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.9229533350943666e-05,
+      "loss": 2.1663,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9229302555264606e-05,
+      "loss": 1.9522,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.922907172640839e-05,
+      "loss": 1.7309,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9228840864375843e-05,
+      "loss": 1.9829,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9228609969167792e-05,
+      "loss": 1.7406,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9228379040785073e-05,
+      "loss": 1.9805,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9228148079228515e-05,
+      "loss": 1.7264,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9227917084498945e-05,
+      "loss": 2.0638,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.92276860565972e-05,
+      "loss": 1.6016,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.92274549955241e-05,
+      "loss": 1.9564,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9227223901280485e-05,
+      "loss": 1.838,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9226992773867182e-05,
+      "loss": 1.6685,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9226761613285017e-05,
+      "loss": 1.706,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9226530419534834e-05,
+      "loss": 1.8744,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9226299192617454e-05,
+      "loss": 2.0391,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.922606793253371e-05,
+      "loss": 1.7639,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.922583663928443e-05,
+      "loss": 1.9809,
+      "step": 6616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9225605312870452e-05,
+      "loss": 1.8857,
+      "step": 6617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9225373953292604e-05,
+      "loss": 2.1329,
+      "step": 6618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.922514256055172e-05,
+      "loss": 1.6394,
+      "step": 6619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9224911134648626e-05,
+      "loss": 2.0367,
+      "step": 6620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9224679675584163e-05,
+      "loss": 1.8421,
+      "step": 6621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9224448183359156e-05,
+      "loss": 1.7722,
+      "step": 6622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.922421665797444e-05,
+      "loss": 2.0105,
+      "step": 6623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9223985099430846e-05,
+      "loss": 1.5404,
+      "step": 6624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9223753507729206e-05,
+      "loss": 1.9563,
+      "step": 6625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9223521882870356e-05,
+      "loss": 2.3197,
+      "step": 6626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9223290224855123e-05,
+      "loss": 1.7962,
+      "step": 6627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9223058533684344e-05,
+      "loss": 1.9741,
+      "step": 6628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9222826809358852e-05,
+      "loss": 1.4611,
+      "step": 6629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9222595051879478e-05,
+      "loss": 2.1214,
+      "step": 6630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9222363261247055e-05,
+      "loss": 2.0608,
+      "step": 6631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.922213143746242e-05,
+      "loss": 1.7326,
+      "step": 6632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.92218995805264e-05,
+      "loss": 1.7843,
+      "step": 6633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9221667690439834e-05,
+      "loss": 2.1573,
+      "step": 6634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.9221435767203552e-05,
+      "loss": 1.9324,
+      "step": 6635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9221203810818392e-05,
+      "loss": 1.7888,
+      "step": 6636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.922097182128518e-05,
+      "loss": 2.2066,
+      "step": 6637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.922073979860476e-05,
+      "loss": 1.7059,
+      "step": 6638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.922050774277796e-05,
+      "loss": 2.0056,
+      "step": 6639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9220275653805614e-05,
+      "loss": 1.9163,
+      "step": 6640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9220043531688558e-05,
+      "loss": 1.7689,
+      "step": 6641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9219811376427627e-05,
+      "loss": 1.9373,
+      "step": 6642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9219579188023653e-05,
+      "loss": 2.8267,
+      "step": 6643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9219346966477474e-05,
+      "loss": 2.0233,
+      "step": 6644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9219114711789923e-05,
+      "loss": 1.9998,
+      "step": 6645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9218882423961834e-05,
+      "loss": 1.7955,
+      "step": 6646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9218650102994042e-05,
+      "loss": 1.6072,
+      "step": 6647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9218417748887384e-05,
+      "loss": 2.3418,
+      "step": 6648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9218185361642693e-05,
+      "loss": 1.9125,
+      "step": 6649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9217952941260806e-05,
+      "loss": 1.8087,
+      "step": 6650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9217720487742556e-05,
+      "loss": 1.8145,
+      "step": 6651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9217488001088784e-05,
+      "loss": 1.9882,
+      "step": 6652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9217255481300323e-05,
+      "loss": 1.9707,
+      "step": 6653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9217022928378006e-05,
+      "loss": 1.7796,
+      "step": 6654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9216790342322675e-05,
+      "loss": 1.6047,
+      "step": 6655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.921655772313516e-05,
+      "loss": 1.986,
+      "step": 6656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.92163250708163e-05,
+      "loss": 1.8431,
+      "step": 6657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9216092385366928e-05,
+      "loss": 2.1024,
+      "step": 6658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9215859666787888e-05,
+      "loss": 1.7901,
+      "step": 6659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.25,
+      "learning_rate": 1.921562691508001e-05,
+      "loss": 2.2063,
+      "step": 6660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9215394130244132e-05,
+      "loss": 1.8372,
+      "step": 6661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9215161312281094e-05,
+      "loss": 1.9711,
+      "step": 6662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9214928461191728e-05,
+      "loss": 1.7058,
+      "step": 6663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9214695576976876e-05,
+      "loss": 1.8345,
+      "step": 6664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.921446265963737e-05,
+      "loss": 1.9795,
+      "step": 6665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.921422970917405e-05,
+      "loss": 1.7406,
+      "step": 6666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9213996725587754e-05,
+      "loss": 1.8569,
+      "step": 6667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.921376370887932e-05,
+      "loss": 2.1034,
+      "step": 6668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.921353065904958e-05,
+      "loss": 2.181,
+      "step": 6669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9213297576099382e-05,
+      "loss": 2.211,
+      "step": 6670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9213064460029552e-05,
+      "loss": 1.7812,
+      "step": 6671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.921283131084094e-05,
+      "loss": 1.8951,
+      "step": 6672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9212598128534377e-05,
+      "loss": 2.176,
+      "step": 6673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9212364913110703e-05,
+      "loss": 1.7248,
+      "step": 6674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.921213166457075e-05,
+      "loss": 2.0868,
+      "step": 6675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9211898382915368e-05,
+      "loss": 1.6301,
+      "step": 6676
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9211665068145388e-05,
+      "loss": 1.8176,
+      "step": 6677
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9211431720261648e-05,
+      "loss": 2.1305,
+      "step": 6678
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9211198339264994e-05,
+      "loss": 1.8219,
+      "step": 6679
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.921096492515626e-05,
+      "loss": 1.7829,
+      "step": 6680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9210731477936284e-05,
+      "loss": 1.8932,
+      "step": 6681
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9210497997605905e-05,
+      "loss": 2.0435,
+      "step": 6682
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9210264484165967e-05,
+      "loss": 2.0036,
+      "step": 6683
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9210030937617303e-05,
+      "loss": 1.8566,
+      "step": 6684
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9209797357960757e-05,
+      "loss": 1.8169,
+      "step": 6685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.920956374519717e-05,
+      "loss": 2.0798,
+      "step": 6686
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9209330099327377e-05,
+      "loss": 1.7702,
+      "step": 6687
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.920909642035222e-05,
+      "loss": 1.7157,
+      "step": 6688
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9208862708272543e-05,
+      "loss": 1.8979,
+      "step": 6689
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.920862896308918e-05,
+      "loss": 2.0195,
+      "step": 6690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9208395184802972e-05,
+      "loss": 1.7067,
+      "step": 6691
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9208161373414763e-05,
+      "loss": 1.8363,
+      "step": 6692
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.920792752892539e-05,
+      "loss": 1.9363,
+      "step": 6693
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9207693651335695e-05,
+      "loss": 2.0283,
+      "step": 6694
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.920745974064652e-05,
+      "loss": 2.1019,
+      "step": 6695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9207225796858706e-05,
+      "loss": 1.8204,
+      "step": 6696
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.920699181997309e-05,
+      "loss": 2.0165,
+      "step": 6697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.920675780999052e-05,
+      "loss": 1.9749,
+      "step": 6698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9206523766911826e-05,
+      "loss": 2.0993,
+      "step": 6699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.920628969073786e-05,
+      "loss": 2.1231,
+      "step": 6700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9206055581469462e-05,
+      "loss": 2.0374,
+      "step": 6701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.920582143910747e-05,
+      "loss": 2.1592,
+      "step": 6702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9205587263652726e-05,
+      "loss": 1.7616,
+      "step": 6703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9205353055106072e-05,
+      "loss": 1.5363,
+      "step": 6704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9205118813468355e-05,
+      "loss": 1.7672,
+      "step": 6705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.9204884538740408e-05,
+      "loss": 2.0658,
+      "step": 6706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9204650230923078e-05,
+      "loss": 2.0623,
+      "step": 6707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9204415890017213e-05,
+      "loss": 2.2088,
+      "step": 6708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9204181516023642e-05,
+      "loss": 1.7778,
+      "step": 6709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9203947108943222e-05,
+      "loss": 1.8293,
+      "step": 6710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9203712668776785e-05,
+      "loss": 1.5145,
+      "step": 6711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.920347819552518e-05,
+      "loss": 1.7622,
+      "step": 6712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9203243689189242e-05,
+      "loss": 1.8963,
+      "step": 6713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9203009149769825e-05,
+      "loss": 1.8489,
+      "step": 6714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9202774577267767e-05,
+      "loss": 2.045,
+      "step": 6715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9202539971683906e-05,
+      "loss": 1.6148,
+      "step": 6716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9202305333019095e-05,
+      "loss": 1.739,
+      "step": 6717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.920207066127417e-05,
+      "loss": 2.0508,
+      "step": 6718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9201835956449975e-05,
+      "loss": 1.882,
+      "step": 6719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.920160121854736e-05,
+      "loss": 1.8568,
+      "step": 6720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9201366447567164e-05,
+      "loss": 1.7769,
+      "step": 6721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9201131643510232e-05,
+      "loss": 2.0993,
+      "step": 6722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9200896806377406e-05,
+      "loss": 2.0125,
+      "step": 6723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9200661936169532e-05,
+      "loss": 2.0198,
+      "step": 6724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9200427032887457e-05,
+      "loss": 1.9185,
+      "step": 6725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.920019209653202e-05,
+      "loss": 1.7908,
+      "step": 6726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9199957127104068e-05,
+      "loss": 1.7242,
+      "step": 6727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9199722124604444e-05,
+      "loss": 1.7742,
+      "step": 6728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9199487089034e-05,
+      "loss": 1.5788,
+      "step": 6729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.919925202039357e-05,
+      "loss": 1.8514,
+      "step": 6730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9199016918684007e-05,
+      "loss": 1.6437,
+      "step": 6731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9198781783906155e-05,
+      "loss": 1.7075,
+      "step": 6732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9198546616060857e-05,
+      "loss": 1.6454,
+      "step": 6733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.919831141514896e-05,
+      "loss": 1.8545,
+      "step": 6734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9198076181171307e-05,
+      "loss": 1.5148,
+      "step": 6735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.919784091412875e-05,
+      "loss": 1.7958,
+      "step": 6736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9197605614022126e-05,
+      "loss": 2.0382,
+      "step": 6737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9197370280852285e-05,
+      "loss": 1.932,
+      "step": 6738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9197134914620074e-05,
+      "loss": 1.9943,
+      "step": 6739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.9196899515326337e-05,
+      "loss": 2.2127,
+      "step": 6740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9196664082971922e-05,
+      "loss": 1.9602,
+      "step": 6741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9196428617557673e-05,
+      "loss": 2.1092,
+      "step": 6742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9196193119084442e-05,
+      "loss": 1.9276,
+      "step": 6743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9195957587553067e-05,
+      "loss": 2.0947,
+      "step": 6744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.91957220229644e-05,
+      "loss": 1.9126,
+      "step": 6745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.919548642531929e-05,
+      "loss": 1.946,
+      "step": 6746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9195250794618578e-05,
+      "loss": 1.921,
+      "step": 6747
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9195015130863117e-05,
+      "loss": 1.9993,
+      "step": 6748
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.919477943405375e-05,
+      "loss": 1.6125,
+      "step": 6749
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9194543704191325e-05,
+      "loss": 1.8572,
+      "step": 6750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.875,
+      "learning_rate": 1.919430794127669e-05,
+      "loss": 2.4487,
+      "step": 6751
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.919407214531069e-05,
+      "loss": 1.9415,
+      "step": 6752
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9193836316294175e-05,
+      "loss": 1.6972,
+      "step": 6753
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9193600454227996e-05,
+      "loss": 2.0991,
+      "step": 6754
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9193364559112997e-05,
+      "loss": 2.4837,
+      "step": 6755
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9193128630950025e-05,
+      "loss": 2.1403,
+      "step": 6756
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.919289266973993e-05,
+      "loss": 2.0499,
+      "step": 6757
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.25,
+      "learning_rate": 1.919265667548356e-05,
+      "loss": 2.3411,
+      "step": 6758
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9192420648181765e-05,
+      "loss": 1.9643,
+      "step": 6759
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9192184587835392e-05,
+      "loss": 1.7401,
+      "step": 6760
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9191948494445288e-05,
+      "loss": 1.6136,
+      "step": 6761
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9191712368012303e-05,
+      "loss": 1.4887,
+      "step": 6762
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9191476208537284e-05,
+      "loss": 1.7767,
+      "step": 6763
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9191240016021086e-05,
+      "loss": 2.2876,
+      "step": 6764
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9191003790464548e-05,
+      "loss": 1.7815,
+      "step": 6765
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.919076753186853e-05,
+      "loss": 1.912,
+      "step": 6766
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9190531240233872e-05,
+      "loss": 2.0169,
+      "step": 6767
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.919029491556143e-05,
+      "loss": 1.8516,
+      "step": 6768
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9190058557852053e-05,
+      "loss": 1.8577,
+      "step": 6769
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9189822167106587e-05,
+      "loss": 1.9014,
+      "step": 6770
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.918958574332588e-05,
+      "loss": 1.9927,
+      "step": 6771
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9189349286510795e-05,
+      "loss": 1.6044,
+      "step": 6772
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9189112796662166e-05,
+      "loss": 1.6214,
+      "step": 6773
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.918887627378085e-05,
+      "loss": 1.6962,
+      "step": 6774
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9188639717867695e-05,
+      "loss": 2.1724,
+      "step": 6775
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9188403128923557e-05,
+      "loss": 1.7977,
+      "step": 6776
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.918816650694928e-05,
+      "loss": 1.9497,
+      "step": 6777
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.918792985194572e-05,
+      "loss": 2.07,
+      "step": 6778
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9187693163913727e-05,
+      "loss": 2.0203,
+      "step": 6779
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9187456442854143e-05,
+      "loss": 1.9063,
+      "step": 6780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.918721968876783e-05,
+      "loss": 1.7206,
+      "step": 6781
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9186982901655634e-05,
+      "loss": 1.7818,
+      "step": 6782
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.918674608151841e-05,
+      "loss": 2.024,
+      "step": 6783
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9186509228357004e-05,
+      "loss": 2.3038,
+      "step": 6784
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.918627234217227e-05,
+      "loss": 1.7492,
+      "step": 6785
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.918603542296506e-05,
+      "loss": 2.0526,
+      "step": 6786
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9185798470736222e-05,
+      "loss": 1.7948,
+      "step": 6787
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.918556148548662e-05,
+      "loss": 1.8615,
+      "step": 6788
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9185324467217085e-05,
+      "loss": 1.6862,
+      "step": 6789
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9185087415928486e-05,
+      "loss": 1.5527,
+      "step": 6790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.918485033162167e-05,
+      "loss": 2.048,
+      "step": 6791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.918461321429749e-05,
+      "loss": 2.2816,
+      "step": 6792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9184376063956795e-05,
+      "loss": 2.1527,
+      "step": 6793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9184138880600445e-05,
+      "loss": 1.9298,
+      "step": 6794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9183901664229282e-05,
+      "loss": 1.5843,
+      "step": 6795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9183664414844166e-05,
+      "loss": 1.8602,
+      "step": 6796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.918342713244595e-05,
+      "loss": 2.2193,
+      "step": 6797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9183189817035485e-05,
+      "loss": 1.8279,
+      "step": 6798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9182952468613624e-05,
+      "loss": 2.0018,
+      "step": 6799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.918271508718122e-05,
+      "loss": 1.8501,
+      "step": 6800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.918247767273913e-05,
+      "loss": 1.973,
+      "step": 6801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.91822402252882e-05,
+      "loss": 1.8808,
+      "step": 6802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.918200274482929e-05,
+      "loss": 1.9946,
+      "step": 6803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9181765231363252e-05,
+      "loss": 2.1289,
+      "step": 6804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.918152768489094e-05,
+      "loss": 1.8816,
+      "step": 6805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9181290105413204e-05,
+      "loss": 1.6536,
+      "step": 6806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9181052492930905e-05,
+      "loss": 1.8331,
+      "step": 6807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.918081484744489e-05,
+      "loss": 1.9277,
+      "step": 6808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.918057716895602e-05,
+      "loss": 1.9753,
+      "step": 6809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9180339457465144e-05,
+      "loss": 1.901,
+      "step": 6810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9180101712973123e-05,
+      "loss": 2.2068,
+      "step": 6811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9179863935480803e-05,
+      "loss": 1.4908,
+      "step": 6812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9179626124989045e-05,
+      "loss": 1.7301,
+      "step": 6813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.91793882814987e-05,
+      "loss": 1.84,
+      "step": 6814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9179150405010624e-05,
+      "loss": 2.1245,
+      "step": 6815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.917891249552568e-05,
+      "loss": 2.1279,
+      "step": 6816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.917867455304471e-05,
+      "loss": 1.1114,
+      "step": 6817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9178436577568576e-05,
+      "loss": 2.0494,
+      "step": 6818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9178198569098137e-05,
+      "loss": 1.8337,
+      "step": 6819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.917796052763424e-05,
+      "loss": 1.9029,
+      "step": 6820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.917772245317775e-05,
+      "loss": 1.7948,
+      "step": 6821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9177484345729515e-05,
+      "loss": 1.7546,
+      "step": 6822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9177246205290397e-05,
+      "loss": 2.0669,
+      "step": 6823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9177008031861244e-05,
+      "loss": 1.7908,
+      "step": 6824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.917676982544292e-05,
+      "loss": 1.887,
+      "step": 6825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9176531586036282e-05,
+      "loss": 1.8891,
+      "step": 6826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.917629331364218e-05,
+      "loss": 1.8499,
+      "step": 6827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9176055008261473e-05,
+      "loss": 1.8888,
+      "step": 6828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.917581666989502e-05,
+      "loss": 1.8051,
+      "step": 6829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9175578298543677e-05,
+      "loss": 2.0773,
+      "step": 6830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.91753398942083e-05,
+      "loss": 1.7863,
+      "step": 6831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.25,
+      "learning_rate": 1.9175101456889743e-05,
+      "loss": 1.7227,
+      "step": 6832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.917486298658887e-05,
+      "loss": 1.8143,
+      "step": 6833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.917462448330653e-05,
+      "loss": 2.1201,
+      "step": 6834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9174385947043587e-05,
+      "loss": 1.9055,
+      "step": 6835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9174147377800894e-05,
+      "loss": 1.9192,
+      "step": 6836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9173908775579316e-05,
+      "loss": 1.8668,
+      "step": 6837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9173670140379702e-05,
+      "loss": 1.8522,
+      "step": 6838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9173431472202914e-05,
+      "loss": 1.7874,
+      "step": 6839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.917319277104981e-05,
+      "loss": 1.8485,
+      "step": 6840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9172954036921246e-05,
+      "loss": 1.908,
+      "step": 6841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.917271526981808e-05,
+      "loss": 1.63,
+      "step": 6842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9172476469741177e-05,
+      "loss": 1.5639,
+      "step": 6843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9172237636691387e-05,
+      "loss": 2.0403,
+      "step": 6844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.917199877066957e-05,
+      "loss": 1.955,
+      "step": 6845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.917175987167659e-05,
+      "loss": 1.8078,
+      "step": 6846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9171520939713303e-05,
+      "loss": 2.0122,
+      "step": 6847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.917128197478056e-05,
+      "loss": 2.2362,
+      "step": 6848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9171042976879234e-05,
+      "loss": 1.9168,
+      "step": 6849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9170803946010178e-05,
+      "loss": 1.7297,
+      "step": 6850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.917056488217425e-05,
+      "loss": 1.6906,
+      "step": 6851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9170325785372306e-05,
+      "loss": 2.0448,
+      "step": 6852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9170086655605212e-05,
+      "loss": 1.7861,
+      "step": 6853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9169847492873828e-05,
+      "loss": 1.6733,
+      "step": 6854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9169608297179008e-05,
+      "loss": 2.0403,
+      "step": 6855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.9169369068521613e-05,
+      "loss": 1.8704,
+      "step": 6856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9169129806902507e-05,
+      "loss": 2.0095,
+      "step": 6857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.916889051232255e-05,
+      "loss": 1.7518,
+      "step": 6858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9168651184782593e-05,
+      "loss": 1.4151,
+      "step": 6859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.916841182428351e-05,
+      "loss": 2.4427,
+      "step": 6860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9168172430826152e-05,
+      "loss": 1.9557,
+      "step": 6861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.916793300441138e-05,
+      "loss": 2.1517,
+      "step": 6862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9167693545040064e-05,
+      "loss": 1.8215,
+      "step": 6863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9167454052713052e-05,
+      "loss": 1.9427,
+      "step": 6864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9167214527431214e-05,
+      "loss": 1.9046,
+      "step": 6865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9166974969195403e-05,
+      "loss": 1.62,
+      "step": 6866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9166735378006487e-05,
+      "loss": 2.0018,
+      "step": 6867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.375,
+      "learning_rate": 1.916649575386533e-05,
+      "loss": 2.0211,
+      "step": 6868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9166256096772782e-05,
+      "loss": 1.593,
+      "step": 6869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9166016406729713e-05,
+      "loss": 1.6881,
+      "step": 6870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9165776683736985e-05,
+      "loss": 1.8809,
+      "step": 6871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9165536927795455e-05,
+      "loss": 1.7658,
+      "step": 6872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.916529713890599e-05,
+      "loss": 1.844,
+      "step": 6873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9165057317069446e-05,
+      "loss": 1.8563,
+      "step": 6874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.916481746228669e-05,
+      "loss": 2.1339,
+      "step": 6875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9164577574558582e-05,
+      "loss": 2.2005,
+      "step": 6876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9164337653885986e-05,
+      "loss": 1.5759,
+      "step": 6877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9164097700269764e-05,
+      "loss": 1.9241,
+      "step": 6878
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9163857713710775e-05,
+      "loss": 2.2982,
+      "step": 6879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9163617694209888e-05,
+      "loss": 1.8612,
+      "step": 6880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.916337764176796e-05,
+      "loss": 1.9616,
+      "step": 6881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.916313755638586e-05,
+      "loss": 1.8906,
+      "step": 6882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9162897438064445e-05,
+      "loss": 2.099,
+      "step": 6883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.916265728680458e-05,
+      "loss": 1.8072,
+      "step": 6884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.916241710260713e-05,
+      "loss": 1.8189,
+      "step": 6885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9162176885472958e-05,
+      "loss": 1.7774,
+      "step": 6886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.9161936635402926e-05,
+      "loss": 1.7477,
+      "step": 6887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.91616963523979e-05,
+      "loss": 1.6292,
+      "step": 6888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.916145603645874e-05,
+      "loss": 1.8947,
+      "step": 6889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9161215687586313e-05,
+      "loss": 1.789,
+      "step": 6890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9160975305781482e-05,
+      "loss": 1.8298,
+      "step": 6891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9160734891045112e-05,
+      "loss": 1.5109,
+      "step": 6892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9160494443378066e-05,
+      "loss": 1.6634,
+      "step": 6893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.916025396278121e-05,
+      "loss": 1.9916,
+      "step": 6894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9160013449255404e-05,
+      "loss": 2.3101,
+      "step": 6895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9159772902801518e-05,
+      "loss": 1.8925,
+      "step": 6896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9159532323420418e-05,
+      "loss": 1.8865,
+      "step": 6897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9159291711112962e-05,
+      "loss": 1.9417,
+      "step": 6898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9159051065880017e-05,
+      "loss": 1.6587,
+      "step": 6899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.915881038772245e-05,
+      "loss": 1.6128,
+      "step": 6900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9158569676641128e-05,
+      "loss": 1.6641,
+      "step": 6901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9158328932636915e-05,
+      "loss": 1.9221,
+      "step": 6902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9158088155710672e-05,
+      "loss": 1.7963,
+      "step": 6903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.915784734586327e-05,
+      "loss": 2.1145,
+      "step": 6904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9157606503095572e-05,
+      "loss": 2.0379,
+      "step": 6905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9157365627408442e-05,
+      "loss": 1.8324,
+      "step": 6906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.915712471880275e-05,
+      "loss": 2.1563,
+      "step": 6907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.915688377727936e-05,
+      "loss": 2.1007,
+      "step": 6908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.915664280283914e-05,
+      "loss": 2.2798,
+      "step": 6909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9156401795482954e-05,
+      "loss": 1.7371,
+      "step": 6910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9156160755211667e-05,
+      "loss": 1.8722,
+      "step": 6911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.915591968202615e-05,
+      "loss": 1.9898,
+      "step": 6912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9155678575927266e-05,
+      "loss": 2.0409,
+      "step": 6913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9155437436915885e-05,
+      "loss": 1.7551,
+      "step": 6914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9155196264992867e-05,
+      "loss": 1.9364,
+      "step": 6915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9154955060159088e-05,
+      "loss": 1.6216,
+      "step": 6916
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.915471382241541e-05,
+      "loss": 2.1463,
+      "step": 6917
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9154472551762694e-05,
+      "loss": 1.8748,
+      "step": 6918
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.915423124820182e-05,
+      "loss": 2.3262,
+      "step": 6919
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.915398991173365e-05,
+      "loss": 1.9161,
+      "step": 6920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.9153748542359052e-05,
+      "loss": 2.0592,
+      "step": 6921
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.915350714007889e-05,
+      "loss": 1.8502,
+      "step": 6922
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.915326570489404e-05,
+      "loss": 2.0998,
+      "step": 6923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9153024236805357e-05,
+      "loss": 1.8833,
+      "step": 6924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9152782735813718e-05,
+      "loss": 1.6875,
+      "step": 6925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.9152541201919992e-05,
+      "loss": 1.94,
+      "step": 6926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9152299635125043e-05,
+      "loss": 1.9565,
+      "step": 6927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9152058035429742e-05,
+      "loss": 1.6531,
+      "step": 6928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9151816402834957e-05,
+      "loss": 2.1461,
+      "step": 6929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9151574737341557e-05,
+      "loss": 2.2188,
+      "step": 6930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.915133303895041e-05,
+      "loss": 1.8328,
+      "step": 6931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9151091307662386e-05,
+      "loss": 2.1203,
+      "step": 6932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.915084954347835e-05,
+      "loss": 1.586,
+      "step": 6933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9150607746399177e-05,
+      "loss": 1.7945,
+      "step": 6934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.915036591642573e-05,
+      "loss": 1.9271,
+      "step": 6935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9150124053558886e-05,
+      "loss": 1.8173,
+      "step": 6936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9149882157799506e-05,
+      "loss": 1.9612,
+      "step": 6937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9149640229148463e-05,
+      "loss": 1.8588,
+      "step": 6938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.914939826760663e-05,
+      "loss": 2.1652,
+      "step": 6939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.914915627317487e-05,
+      "loss": 1.9239,
+      "step": 6940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9148914245854065e-05,
+      "loss": 2.1673,
+      "step": 6941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.625,
+      "learning_rate": 1.914867218564507e-05,
+      "loss": 1.962,
+      "step": 6942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9148430092548764e-05,
+      "loss": 1.8639,
+      "step": 6943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9148187966566015e-05,
+      "loss": 1.9538,
+      "step": 6944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9147945807697696e-05,
+      "loss": 1.6683,
+      "step": 6945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9147703615944674e-05,
+      "loss": 1.9498,
+      "step": 6946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9147461391307818e-05,
+      "loss": 1.7637,
+      "step": 6947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9147219133788004e-05,
+      "loss": 1.7549,
+      "step": 6948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9146976843386097e-05,
+      "loss": 1.7273,
+      "step": 6949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9146734520102977e-05,
+      "loss": 2.2424,
+      "step": 6950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.914649216393951e-05,
+      "loss": 2.0985,
+      "step": 6951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9146249774896562e-05,
+      "loss": 2.1355,
+      "step": 6952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.914600735297501e-05,
+      "loss": 1.9017,
+      "step": 6953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9145764898175726e-05,
+      "loss": 1.4215,
+      "step": 6954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9145522410499578e-05,
+      "loss": 1.77,
+      "step": 6955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9145279889947445e-05,
+      "loss": 1.895,
+      "step": 6956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9145037336520188e-05,
+      "loss": 1.8836,
+      "step": 6957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9144794750218688e-05,
+      "loss": 1.9166,
+      "step": 6958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.914455213104381e-05,
+      "loss": 1.7355,
+      "step": 6959
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9144309478996434e-05,
+      "loss": 1.6113,
+      "step": 6960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9144066794077425e-05,
+      "loss": 2.0477,
+      "step": 6961
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9143824076287657e-05,
+      "loss": 1.6441,
+      "step": 6962
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.914358132562801e-05,
+      "loss": 1.7213,
+      "step": 6963
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9143338542099346e-05,
+      "loss": 2.1434,
+      "step": 6964
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9143095725702545e-05,
+      "loss": 1.856,
+      "step": 6965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.914285287643847e-05,
+      "loss": 1.994,
+      "step": 6966
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.914260999430801e-05,
+      "loss": 2.1281,
+      "step": 6967
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9142367079312023e-05,
+      "loss": 2.0834,
+      "step": 6968
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.9142124131451392e-05,
+      "loss": 1.7924,
+      "step": 6969
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9141881150726984e-05,
+      "loss": 2.0208,
+      "step": 6970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9141638137139676e-05,
+      "loss": 1.8201,
+      "step": 6971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.9141395090690343e-05,
+      "loss": 1.9328,
+      "step": 6972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9141152011379855e-05,
+      "loss": 2.2205,
+      "step": 6973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.914090889920909e-05,
+      "loss": 1.958,
+      "step": 6974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9140665754178914e-05,
+      "loss": 1.7711,
+      "step": 6975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9140422576290213e-05,
+      "loss": 1.944,
+      "step": 6976
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9140179365543847e-05,
+      "loss": 1.8253,
+      "step": 6977
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9139936121940704e-05,
+      "loss": 2.011,
+      "step": 6978
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.913969284548165e-05,
+      "loss": 1.6349,
+      "step": 6979
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9139449536167562e-05,
+      "loss": 1.9424,
+      "step": 6980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.913920619399931e-05,
+      "loss": 1.9054,
+      "step": 6981
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.913896281897778e-05,
+      "loss": 1.7937,
+      "step": 6982
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9138719411103835e-05,
+      "loss": 1.7765,
+      "step": 6983
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9138475970378357e-05,
+      "loss": 1.9002,
+      "step": 6984
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9138232496802218e-05,
+      "loss": 1.6091,
+      "step": 6985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9137988990376298e-05,
+      "loss": 1.6529,
+      "step": 6986
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9137745451101465e-05,
+      "loss": 1.8482,
+      "step": 6987
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.91375018789786e-05,
+      "loss": 1.9056,
+      "step": 6988
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.913725827400857e-05,
+      "loss": 1.9681,
+      "step": 6989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9137014636192266e-05,
+      "loss": 1.8724,
+      "step": 6990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9136770965530553e-05,
+      "loss": 1.8489,
+      "step": 6991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.913652726202431e-05,
+      "loss": 1.9046,
+      "step": 6992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.913628352567441e-05,
+      "loss": 1.754,
+      "step": 6993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.913603975648173e-05,
+      "loss": 1.7029,
+      "step": 6994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.913579595444715e-05,
+      "loss": 1.7037,
+      "step": 6995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9135552119571544e-05,
+      "loss": 1.9413,
+      "step": 6996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.913530825185579e-05,
+      "loss": 1.753,
+      "step": 6997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.913506435130076e-05,
+      "loss": 1.7557,
+      "step": 6998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9134820417907337e-05,
+      "loss": 1.9528,
+      "step": 6999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.913457645167639e-05,
+      "loss": 2.4186,
+      "step": 7000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9134332452608804e-05,
+      "loss": 1.9251,
+      "step": 7001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9134088420705457e-05,
+      "loss": 1.9284,
+      "step": 7002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9133844355967216e-05,
+      "loss": 1.8865,
+      "step": 7003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.913360025839497e-05,
+      "loss": 1.8551,
+      "step": 7004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9133356127989587e-05,
+      "loss": 2.2982,
+      "step": 7005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.913311196475195e-05,
+      "loss": 1.999,
+      "step": 7006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9132867768682936e-05,
+      "loss": 1.9233,
+      "step": 7007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9132623539783425e-05,
+      "loss": 1.7181,
+      "step": 7008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.913237927805429e-05,
+      "loss": 1.9982,
+      "step": 7009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.913213498349641e-05,
+      "loss": 1.8783,
+      "step": 7010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9131890656110662e-05,
+      "loss": 1.7731,
+      "step": 7011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.913164629589793e-05,
+      "loss": 2.0062,
+      "step": 7012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9131401902859087e-05,
+      "loss": 1.874,
+      "step": 7013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.913115747699502e-05,
+      "loss": 1.9355,
+      "step": 7014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9130913018306593e-05,
+      "loss": 1.8251,
+      "step": 7015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9130668526794698e-05,
+      "loss": 1.826,
+      "step": 7016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.913042400246021e-05,
+      "loss": 1.7031,
+      "step": 7017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9130179445304002e-05,
+      "loss": 2.0618,
+      "step": 7018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9129934855326963e-05,
+      "loss": 1.8447,
+      "step": 7019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9129690232529965e-05,
+      "loss": 1.7103,
+      "step": 7020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9129445576913886e-05,
+      "loss": 2.0374,
+      "step": 7021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9129200888479614e-05,
+      "loss": 1.8385,
+      "step": 7022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9128956167228026e-05,
+      "loss": 1.9286,
+      "step": 7023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9128711413159994e-05,
+      "loss": 1.6234,
+      "step": 7024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9128466626276407e-05,
+      "loss": 1.9436,
+      "step": 7025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.912822180657814e-05,
+      "loss": 2.0143,
+      "step": 7026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9127976954066075e-05,
+      "loss": 1.9633,
+      "step": 7027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9127732068741092e-05,
+      "loss": 2.1196,
+      "step": 7028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9127487150604068e-05,
+      "loss": 1.8411,
+      "step": 7029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9127242199655888e-05,
+      "loss": 1.5951,
+      "step": 7030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9126997215897433e-05,
+      "loss": 1.9161,
+      "step": 7031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.912675219932958e-05,
+      "loss": 1.9193,
+      "step": 7032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.912650714995321e-05,
+      "loss": 1.6616,
+      "step": 7033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9126262067769205e-05,
+      "loss": 1.9965,
+      "step": 7034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9126016952778448e-05,
+      "loss": 2.1326,
+      "step": 7035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9125771804981817e-05,
+      "loss": 2.1837,
+      "step": 7036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9125526624380196e-05,
+      "loss": 1.7412,
+      "step": 7037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9125281410974463e-05,
+      "loss": 1.6622,
+      "step": 7038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9125036164765502e-05,
+      "loss": 2.0467,
+      "step": 7039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9124790885754194e-05,
+      "loss": 2.0459,
+      "step": 7040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.912454557394142e-05,
+      "loss": 2.0669,
+      "step": 7041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9124300229328064e-05,
+      "loss": 1.9653,
+      "step": 7042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9124054851915008e-05,
+      "loss": 1.6071,
+      "step": 7043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.912380944170313e-05,
+      "loss": 1.954,
+      "step": 7044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9123563998693315e-05,
+      "loss": 1.8695,
+      "step": 7045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9123318522886445e-05,
+      "loss": 1.9833,
+      "step": 7046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9123073014283404e-05,
+      "loss": 1.8376,
+      "step": 7047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.912282747288507e-05,
+      "loss": 2.0341,
+      "step": 7048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.912258189869233e-05,
+      "loss": 2.1016,
+      "step": 7049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.912233629170607e-05,
+      "loss": 1.9431,
+      "step": 7050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9122090651927162e-05,
+      "loss": 1.9772,
+      "step": 7051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.91218449793565e-05,
+      "loss": 1.7831,
+      "step": 7052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.912159927399496e-05,
+      "loss": 1.5742,
+      "step": 7053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9121353535843428e-05,
+      "loss": 1.7532,
+      "step": 7054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9121107764902788e-05,
+      "loss": 2.1117,
+      "step": 7055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9120861961173922e-05,
+      "loss": 1.7363,
+      "step": 7056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9120616124657715e-05,
+      "loss": 1.9993,
+      "step": 7057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.912037025535505e-05,
+      "loss": 1.8354,
+      "step": 7058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.912012435326681e-05,
+      "loss": 2.3483,
+      "step": 7059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.911987841839388e-05,
+      "loss": 1.756,
+      "step": 7060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9119632450737144e-05,
+      "loss": 1.9168,
+      "step": 7061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9119386450297487e-05,
+      "loss": 1.8674,
+      "step": 7062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9119140417075788e-05,
+      "loss": 2.1754,
+      "step": 7063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9118894351072942e-05,
+      "loss": 1.6485,
+      "step": 7064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9118648252289825e-05,
+      "loss": 2.0601,
+      "step": 7065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9118402120727323e-05,
+      "loss": 1.9184,
+      "step": 7066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9118155956386323e-05,
+      "loss": 1.5464,
+      "step": 7067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9117909759267706e-05,
+      "loss": 1.7544,
+      "step": 7068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9117663529372363e-05,
+      "loss": 2.1153,
+      "step": 7069
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9117417266701175e-05,
+      "loss": 2.0076,
+      "step": 7070
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9117170971255027e-05,
+      "loss": 1.7339,
+      "step": 7071
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.911692464303481e-05,
+      "loss": 1.8162,
+      "step": 7072
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.91166782820414e-05,
+      "loss": 1.5831,
+      "step": 7073
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.911643188827569e-05,
+      "loss": 1.879,
+      "step": 7074
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9116185461738565e-05,
+      "loss": 1.9131,
+      "step": 7075
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9115939002430904e-05,
+      "loss": 1.5855,
+      "step": 7076
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.91156925103536e-05,
+      "loss": 1.7775,
+      "step": 7077
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9115445985507536e-05,
+      "loss": 1.5252,
+      "step": 7078
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.91151994278936e-05,
+      "loss": 2.0831,
+      "step": 7079
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9114952837512684e-05,
+      "loss": 1.7471,
+      "step": 7080
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9114706214365663e-05,
+      "loss": 1.8382,
+      "step": 7081
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.911445955845343e-05,
+      "loss": 2.0455,
+      "step": 7082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9114212869776866e-05,
+      "loss": 1.7931,
+      "step": 7083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9113966148336866e-05,
+      "loss": 1.9677,
+      "step": 7084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9113719394134315e-05,
+      "loss": 2.5052,
+      "step": 7085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9113472607170096e-05,
+      "loss": 1.8538,
+      "step": 7086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9113225787445098e-05,
+      "loss": 1.9239,
+      "step": 7087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9112978934960207e-05,
+      "loss": 1.9194,
+      "step": 7088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9112732049716313e-05,
+      "loss": 1.9841,
+      "step": 7089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9112485131714304e-05,
+      "loss": 2.2381,
+      "step": 7090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9112238180955064e-05,
+      "loss": 1.5309,
+      "step": 7091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9111991197439485e-05,
+      "loss": 1.9648,
+      "step": 7092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.911174418116845e-05,
+      "loss": 1.9787,
+      "step": 7093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9111497132142852e-05,
+      "loss": 1.8199,
+      "step": 7094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9111250050363574e-05,
+      "loss": 1.9352,
+      "step": 7095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.911100293583151e-05,
+      "loss": 1.6705,
+      "step": 7096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9110755788547545e-05,
+      "loss": 2.1499,
+      "step": 7097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9110508608512563e-05,
+      "loss": 2.0192,
+      "step": 7098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.911026139572746e-05,
+      "loss": 1.7562,
+      "step": 7099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.911001415019312e-05,
+      "loss": 2.1898,
+      "step": 7100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.910976687191044e-05,
+      "loss": 2.0219,
+      "step": 7101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9109519560880297e-05,
+      "loss": 1.7587,
+      "step": 7102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9109272217103584e-05,
+      "loss": 2.0406,
+      "step": 7103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9109024840581197e-05,
+      "loss": 1.9275,
+      "step": 7104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9108777431314018e-05,
+      "loss": 1.9485,
+      "step": 7105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.9108529989302937e-05,
+      "loss": 1.9763,
+      "step": 7106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.910828251454884e-05,
+      "loss": 1.855,
+      "step": 7107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9108035007052626e-05,
+      "loss": 2.0295,
+      "step": 7108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9107787466815183e-05,
+      "loss": 1.5993,
+      "step": 7109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9107539893837396e-05,
+      "loss": 1.7583,
+      "step": 7110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9107292288120155e-05,
+      "loss": 1.8626,
+      "step": 7111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9107044649664356e-05,
+      "loss": 1.5706,
+      "step": 7112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9106796978470882e-05,
+      "loss": 2.058,
+      "step": 7113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9106549274540625e-05,
+      "loss": 1.9181,
+      "step": 7114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.910630153787448e-05,
+      "loss": 1.9309,
+      "step": 7115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.9106053768473336e-05,
+      "loss": 1.789,
+      "step": 7116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.910580596633808e-05,
+      "loss": 1.9428,
+      "step": 7117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9105558131469607e-05,
+      "loss": 2.0222,
+      "step": 7118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.91053102638688e-05,
+      "loss": 2.1796,
+      "step": 7119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9105062363536563e-05,
+      "loss": 2.1447,
+      "step": 7120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9104814430473777e-05,
+      "loss": 1.7821,
+      "step": 7121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9104566464681337e-05,
+      "loss": 1.8486,
+      "step": 7122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9104318466160132e-05,
+      "loss": 1.7373,
+      "step": 7123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9104070434911056e-05,
+      "loss": 1.6699,
+      "step": 7124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9103822370935e-05,
+      "loss": 1.6754,
+      "step": 7125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9103574274232856e-05,
+      "loss": 1.6747,
+      "step": 7126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9103326144805515e-05,
+      "loss": 1.603,
+      "step": 7127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.910307798265387e-05,
+      "loss": 2.1356,
+      "step": 7128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.910282978777881e-05,
+      "loss": 1.8581,
+      "step": 7129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9102581560181235e-05,
+      "loss": 1.8205,
+      "step": 7130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.910233329986203e-05,
+      "loss": 1.7706,
+      "step": 7131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.9102085006822085e-05,
+      "loss": 2.6345,
+      "step": 7132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9101836681062302e-05,
+      "loss": 2.0275,
+      "step": 7133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9101588322583568e-05,
+      "loss": 1.7132,
+      "step": 7134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9101339931386773e-05,
+      "loss": 1.9136,
+      "step": 7135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.910109150747282e-05,
+      "loss": 1.8024,
+      "step": 7136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9100843050842586e-05,
+      "loss": 1.9631,
+      "step": 7137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.910059456149698e-05,
+      "loss": 1.9835,
+      "step": 7138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9100346039436888e-05,
+      "loss": 2.0875,
+      "step": 7139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.91000974846632e-05,
+      "loss": 2.1254,
+      "step": 7140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.909984889717682e-05,
+      "loss": 1.6206,
+      "step": 7141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.909960027697863e-05,
+      "loss": 1.7722,
+      "step": 7142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9099351624069533e-05,
+      "loss": 1.9588,
+      "step": 7143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.909910293845042e-05,
+      "loss": 1.6825,
+      "step": 7144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9098854220122177e-05,
+      "loss": 1.8801,
+      "step": 7145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9098605469085712e-05,
+      "loss": 1.939,
+      "step": 7146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.909835668534191e-05,
+      "loss": 1.6288,
+      "step": 7147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9098107868891668e-05,
+      "loss": 2.2558,
+      "step": 7148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9097859019735875e-05,
+      "loss": 1.6964,
+      "step": 7149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9097610137875435e-05,
+      "loss": 1.6297,
+      "step": 7150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9097361223311235e-05,
+      "loss": 1.8229,
+      "step": 7151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9097112276044175e-05,
+      "loss": 2.0174,
+      "step": 7152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9096863296075148e-05,
+      "loss": 1.7073,
+      "step": 7153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9096614283405048e-05,
+      "loss": 2.0143,
+      "step": 7154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.909636523803477e-05,
+      "loss": 1.8907,
+      "step": 7155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9096116159965213e-05,
+      "loss": 1.9582,
+      "step": 7156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.9095867049197267e-05,
+      "loss": 1.9877,
+      "step": 7157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9095617905731833e-05,
+      "loss": 2.3167,
+      "step": 7158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9095368729569802e-05,
+      "loss": 1.7942,
+      "step": 7159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9095119520712073e-05,
+      "loss": 1.9776,
+      "step": 7160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9094870279159535e-05,
+      "loss": 1.9744,
+      "step": 7161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9094621004913092e-05,
+      "loss": 2.0974,
+      "step": 7162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.909437169797364e-05,
+      "loss": 1.9653,
+      "step": 7163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.909412235834207e-05,
+      "loss": 1.7788,
+      "step": 7164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.909387298601928e-05,
+      "loss": 2.0154,
+      "step": 7165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.909362358100617e-05,
+      "loss": 1.5491,
+      "step": 7166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.909337414330363e-05,
+      "loss": 1.9466,
+      "step": 7167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9093124672912563e-05,
+      "loss": 2.5084,
+      "step": 7168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9092875169833863e-05,
+      "loss": 1.9013,
+      "step": 7169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9092625634068426e-05,
+      "loss": 1.9092,
+      "step": 7170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9092376065617154e-05,
+      "loss": 1.9284,
+      "step": 7171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9092126464480936e-05,
+      "loss": 1.7806,
+      "step": 7172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9091876830660675e-05,
+      "loss": 1.8819,
+      "step": 7173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.9091627164157268e-05,
+      "loss": 1.9798,
+      "step": 7174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9091377464971613e-05,
+      "loss": 1.9221,
+      "step": 7175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.90911277331046e-05,
+      "loss": 1.6358,
+      "step": 7176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.909087796855714e-05,
+      "loss": 1.7855,
+      "step": 7177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.909062817133012e-05,
+      "loss": 1.7229,
+      "step": 7178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9090378341424442e-05,
+      "loss": 2.0465,
+      "step": 7179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9090128478841002e-05,
+      "loss": 1.9987,
+      "step": 7180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9089878583580704e-05,
+      "loss": 1.9041,
+      "step": 7181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9089628655644438e-05,
+      "loss": 1.9873,
+      "step": 7182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.908937869503311e-05,
+      "loss": 1.7726,
+      "step": 7183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9089128701747613e-05,
+      "loss": 1.9772,
+      "step": 7184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9088878675788844e-05,
+      "loss": 1.9235,
+      "step": 7185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.908862861715771e-05,
+      "loss": 2.0199,
+      "step": 7186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9088378525855108e-05,
+      "loss": 1.7783,
+      "step": 7187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9088128401881932e-05,
+      "loss": 1.7542,
+      "step": 7188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.908787824523908e-05,
+      "loss": 1.7941,
+      "step": 7189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.908762805592746e-05,
+      "loss": 1.9784,
+      "step": 7190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.9087377833947965e-05,
+      "loss": 1.952,
+      "step": 7191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9087127579301495e-05,
+      "loss": 1.6852,
+      "step": 7192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9086877291988953e-05,
+      "loss": 2.0685,
+      "step": 7193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9086626972011235e-05,
+      "loss": 1.9253,
+      "step": 7194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.908637661936924e-05,
+      "loss": 2.1915,
+      "step": 7195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.908612623406387e-05,
+      "loss": 1.7568,
+      "step": 7196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9085875816096026e-05,
+      "loss": 2.0184,
+      "step": 7197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.908562536546661e-05,
+      "loss": 1.9655,
+      "step": 7198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9085374882176514e-05,
+      "loss": 2.2461,
+      "step": 7199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.908512436622665e-05,
+      "loss": 2.0752,
+      "step": 7200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9084873817617906e-05,
+      "loss": 2.2426,
+      "step": 7201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.908462323635119e-05,
+      "loss": 2.0862,
+      "step": 7202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9084372622427406e-05,
+      "loss": 1.8793,
+      "step": 7203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9084121975847448e-05,
+      "loss": 1.7402,
+      "step": 7204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.908387129661222e-05,
+      "loss": 1.8789,
+      "step": 7205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9083620584722625e-05,
+      "loss": 2.0354,
+      "step": 7206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9083369840179562e-05,
+      "loss": 1.6512,
+      "step": 7207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9083119062983932e-05,
+      "loss": 1.7808,
+      "step": 7208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9082868253136636e-05,
+      "loss": 1.7423,
+      "step": 7209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9082617410638577e-05,
+      "loss": 1.8166,
+      "step": 7210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9082366535490657e-05,
+      "loss": 1.9656,
+      "step": 7211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9082115627693775e-05,
+      "loss": 2.0176,
+      "step": 7212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9081864687248838e-05,
+      "loss": 1.8416,
+      "step": 7213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9081613714156747e-05,
+      "loss": 1.8032,
+      "step": 7214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9081362708418397e-05,
+      "loss": 1.6883,
+      "step": 7215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.90811116700347e-05,
+      "loss": 1.998,
+      "step": 7216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.908086059900655e-05,
+      "loss": 2.2427,
+      "step": 7217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9080609495334855e-05,
+      "loss": 1.5171,
+      "step": 7218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.908035835902052e-05,
+      "loss": 1.9605,
+      "step": 7219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.9080107190064442e-05,
+      "loss": 1.3154,
+      "step": 7220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.9079855988467528e-05,
+      "loss": 2.1049,
+      "step": 7221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9079604754230673e-05,
+      "loss": 1.9507,
+      "step": 7222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9079353487354793e-05,
+      "loss": 1.9422,
+      "step": 7223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.907910218784078e-05,
+      "loss": 2.1562,
+      "step": 7224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9078850855689547e-05,
+      "loss": 2.2616,
+      "step": 7225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9078599490901984e-05,
+      "loss": 1.8489,
+      "step": 7226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9078348093479008e-05,
+      "loss": 1.7269,
+      "step": 7227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9078096663421516e-05,
+      "loss": 1.703,
+      "step": 7228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9077845200730415e-05,
+      "loss": 1.7772,
+      "step": 7229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.907759370540661e-05,
+      "loss": 1.9036,
+      "step": 7230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9077342177451e-05,
+      "loss": 1.9215,
+      "step": 7231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.907709061686449e-05,
+      "loss": 2.0198,
+      "step": 7232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9076839023647988e-05,
+      "loss": 1.7561,
+      "step": 7233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.907658739780239e-05,
+      "loss": 1.7359,
+      "step": 7234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9076335739328617e-05,
+      "loss": 1.7286,
+      "step": 7235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9076084048227557e-05,
+      "loss": 2.0512,
+      "step": 7236
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.907583232450012e-05,
+      "loss": 1.8642,
+      "step": 7237
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9075580568147217e-05,
+      "loss": 1.8461,
+      "step": 7238
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9075328779169745e-05,
+      "loss": 2.1085,
+      "step": 7239
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9075076957568613e-05,
+      "loss": 1.888,
+      "step": 7240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9074825103344725e-05,
+      "loss": 1.9108,
+      "step": 7241
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9074573216498987e-05,
+      "loss": 1.9986,
+      "step": 7242
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9074321297032305e-05,
+      "loss": 1.7344,
+      "step": 7243
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9074069344945585e-05,
+      "loss": 2.2729,
+      "step": 7244
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.907381736023973e-05,
+      "loss": 1.9819,
+      "step": 7245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9073565342915648e-05,
+      "loss": 1.6448,
+      "step": 7246
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9073313292974247e-05,
+      "loss": 1.58,
+      "step": 7247
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9073061210416427e-05,
+      "loss": 1.9003,
+      "step": 7248
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9072809095243098e-05,
+      "loss": 1.8687,
+      "step": 7249
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.907255694745517e-05,
+      "loss": 1.9992,
+      "step": 7250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9072304767053536e-05,
+      "loss": 1.9902,
+      "step": 7251
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9072052554039123e-05,
+      "loss": 1.8192,
+      "step": 7252
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.907180030841282e-05,
+      "loss": 2.2371,
+      "step": 7253
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9071548030175542e-05,
+      "loss": 1.624,
+      "step": 7254
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9071295719328193e-05,
+      "loss": 1.9572,
+      "step": 7255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9071043375871682e-05,
+      "loss": 2.0216,
+      "step": 7256
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9070790999806916e-05,
+      "loss": 1.9521,
+      "step": 7257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.90705385911348e-05,
+      "loss": 1.9642,
+      "step": 7258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9070286149856247e-05,
+      "loss": 1.6351,
+      "step": 7259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9070033675972156e-05,
+      "loss": 1.9134,
+      "step": 7260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9069781169483442e-05,
+      "loss": 1.8548,
+      "step": 7261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9069528630391007e-05,
+      "loss": 1.9167,
+      "step": 7262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.906927605869576e-05,
+      "loss": 2.261,
+      "step": 7263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.9069023454398613e-05,
+      "loss": 1.7825,
+      "step": 7264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9068770817500473e-05,
+      "loss": 1.9321,
+      "step": 7265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9068518148002244e-05,
+      "loss": 1.7452,
+      "step": 7266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.9068265445904838e-05,
+      "loss": 1.7283,
+      "step": 7267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.906801271120916e-05,
+      "loss": 1.6903,
+      "step": 7268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9067759943916124e-05,
+      "loss": 1.7705,
+      "step": 7269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9067507144026635e-05,
+      "loss": 2.0169,
+      "step": 7270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.90672543115416e-05,
+      "loss": 1.86,
+      "step": 7271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9067001446461934e-05,
+      "loss": 1.7902,
+      "step": 7272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.906674854878854e-05,
+      "loss": 1.9613,
+      "step": 7273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.9066495618522327e-05,
+      "loss": 2.2128,
+      "step": 7274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.906624265566421e-05,
+      "loss": 2.2563,
+      "step": 7275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.906598966021509e-05,
+      "loss": 1.8147,
+      "step": 7276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9065736632175886e-05,
+      "loss": 1.8026,
+      "step": 7277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.90654835715475e-05,
+      "loss": 1.6594,
+      "step": 7278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9065230478330846e-05,
+      "loss": 1.6707,
+      "step": 7279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9064977352526834e-05,
+      "loss": 1.7975,
+      "step": 7280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9064724194136366e-05,
+      "loss": 1.991,
+      "step": 7281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9064471003160365e-05,
+      "loss": 1.9293,
+      "step": 7282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9064217779599732e-05,
+      "loss": 2.1532,
+      "step": 7283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.906396452345538e-05,
+      "loss": 1.8812,
+      "step": 7284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.906371123472822e-05,
+      "loss": 2.6296,
+      "step": 7285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9063457913419158e-05,
+      "loss": 2.181,
+      "step": 7286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9063204559529113e-05,
+      "loss": 2.0206,
+      "step": 7287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.9062951173058985e-05,
+      "loss": 2.1978,
+      "step": 7288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.9062697754009697e-05,
+      "loss": 1.876,
+      "step": 7289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.906244430238215e-05,
+      "loss": 1.949,
+      "step": 7290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.906219081817726e-05,
+      "loss": 1.7834,
+      "step": 7291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9061937301395938e-05,
+      "loss": 1.7199,
+      "step": 7292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9061683752039094e-05,
+      "loss": 1.7164,
+      "step": 7293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.9061430170107637e-05,
+      "loss": 2.0058,
+      "step": 7294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9061176555602484e-05,
+      "loss": 1.9681,
+      "step": 7295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9060922908524547e-05,
+      "loss": 2.1054,
+      "step": 7296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9060669228874734e-05,
+      "loss": 1.8147,
+      "step": 7297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9060415516653956e-05,
+      "loss": 1.6481,
+      "step": 7298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9060161771863124e-05,
+      "loss": 2.0548,
+      "step": 7299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9059907994503158e-05,
+      "loss": 1.9741,
+      "step": 7300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.905965418457496e-05,
+      "loss": 1.967,
+      "step": 7301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9059400342079456e-05,
+      "loss": 1.7214,
+      "step": 7302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9059146467017546e-05,
+      "loss": 1.8385,
+      "step": 7303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9058892559390143e-05,
+      "loss": 1.8831,
+      "step": 7304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.905863861919817e-05,
+      "loss": 1.6339,
+      "step": 7305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9058384646442528e-05,
+      "loss": 1.5962,
+      "step": 7306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.905813064112414e-05,
+      "loss": 1.9155,
+      "step": 7307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9057876603243912e-05,
+      "loss": 1.9944,
+      "step": 7308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.905762253280276e-05,
+      "loss": 1.5501,
+      "step": 7309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.9057368429801598e-05,
+      "loss": 1.9859,
+      "step": 7310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.905711429424134e-05,
+      "loss": 1.9268,
+      "step": 7311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9056860126122895e-05,
+      "loss": 2.1214,
+      "step": 7312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9056605925447182e-05,
+      "loss": 1.7579,
+      "step": 7313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.905635169221511e-05,
+      "loss": 1.8835,
+      "step": 7314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9056097426427594e-05,
+      "loss": 1.5847,
+      "step": 7315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9055843128085553e-05,
+      "loss": 1.8791,
+      "step": 7316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9055588797189898e-05,
+      "loss": 1.9964,
+      "step": 7317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.905533443374154e-05,
+      "loss": 1.8872,
+      "step": 7318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.90550800377414e-05,
+      "loss": 1.9811,
+      "step": 7319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.9054825609190386e-05,
+      "loss": 2.0128,
+      "step": 7320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.9054571148089416e-05,
+      "loss": 1.6543,
+      "step": 7321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9054316654439402e-05,
+      "loss": 1.7812,
+      "step": 7322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9054062128241263e-05,
+      "loss": 1.8055,
+      "step": 7323
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9053807569495913e-05,
+      "loss": 1.9008,
+      "step": 7324
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9053552978204265e-05,
+      "loss": 2.0327,
+      "step": 7325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9053298354367235e-05,
+      "loss": 1.8188,
+      "step": 7326
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.905304369798574e-05,
+      "loss": 1.8107,
+      "step": 7327
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.905278900906069e-05,
+      "loss": 1.9776,
+      "step": 7328
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.905253428759301e-05,
+      "loss": 1.8367,
+      "step": 7329
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9052279533583604e-05,
+      "loss": 1.937,
+      "step": 7330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9052024747033395e-05,
+      "loss": 1.9749,
+      "step": 7331
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.90517699279433e-05,
+      "loss": 1.6626,
+      "step": 7332
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9051515076314233e-05,
+      "loss": 2.0548,
+      "step": 7333
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.905126019214711e-05,
+      "loss": 1.956,
+      "step": 7334
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.9051005275442844e-05,
+      "loss": 2.0174,
+      "step": 7335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.905075032620236e-05,
+      "loss": 2.226,
+      "step": 7336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9050495344426564e-05,
+      "loss": 1.7631,
+      "step": 7337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.9050240330116382e-05,
+      "loss": 1.8905,
+      "step": 7338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.904998528327272e-05,
+      "loss": 1.9615,
+      "step": 7339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9049730203896508e-05,
+      "loss": 1.9334,
+      "step": 7340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.904947509198865e-05,
+      "loss": 1.9626,
+      "step": 7341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9049219947550072e-05,
+      "loss": 1.9347,
+      "step": 7342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.904896477058169e-05,
+      "loss": 1.9604,
+      "step": 7343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.9048709561084417e-05,
+      "loss": 2.001,
+      "step": 7344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9048454319059174e-05,
+      "loss": 1.8246,
+      "step": 7345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.904819904450688e-05,
+      "loss": 2.3093,
+      "step": 7346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9047943737428448e-05,
+      "loss": 1.9625,
+      "step": 7347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.90476883978248e-05,
+      "loss": 1.631,
+      "step": 7348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.904743302569685e-05,
+      "loss": 1.8093,
+      "step": 7349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.904717762104552e-05,
+      "loss": 1.9794,
+      "step": 7350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9046922183871722e-05,
+      "loss": 2.2886,
+      "step": 7351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.904666671417638e-05,
+      "loss": 1.8671,
+      "step": 7352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.904641121196041e-05,
+      "loss": 1.5921,
+      "step": 7353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9046155677224733e-05,
+      "loss": 1.8436,
+      "step": 7354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9045900109970267e-05,
+      "loss": 1.6929,
+      "step": 7355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.9045644510197925e-05,
+      "loss": 1.7188,
+      "step": 7356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9045388877908633e-05,
+      "loss": 1.8414,
+      "step": 7357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9045133213103307e-05,
+      "loss": 1.9159,
+      "step": 7358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9044877515782866e-05,
+      "loss": 1.8754,
+      "step": 7359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.904462178594823e-05,
+      "loss": 1.6981,
+      "step": 7360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9044366023600317e-05,
+      "loss": 1.7708,
+      "step": 7361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.9044110228740048e-05,
+      "loss": 1.7604,
+      "step": 7362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9043854401368337e-05,
+      "loss": 2.2268,
+      "step": 7363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9043598541486116e-05,
+      "loss": 2.0198,
+      "step": 7364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9043342649094293e-05,
+      "loss": 1.9226,
+      "step": 7365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.904308672419379e-05,
+      "loss": 2.0352,
+      "step": 7366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.904283076678553e-05,
+      "loss": 1.7139,
+      "step": 7367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9042574776870432e-05,
+      "loss": 2.0426,
+      "step": 7368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.904231875444942e-05,
+      "loss": 1.8601,
+      "step": 7369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9042062699523404e-05,
+      "loss": 1.7926,
+      "step": 7370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9041806612093317e-05,
+      "loss": 1.7828,
+      "step": 7371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.904155049216007e-05,
+      "loss": 2.5627,
+      "step": 7372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9041294339724587e-05,
+      "loss": 2.1777,
+      "step": 7373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.904103815478779e-05,
+      "loss": 1.8086,
+      "step": 7374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9040781937350597e-05,
+      "loss": 1.9733,
+      "step": 7375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9040525687413935e-05,
+      "loss": 1.8412,
+      "step": 7376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.904026940497872e-05,
+      "loss": 1.6494,
+      "step": 7377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9040013090045873e-05,
+      "loss": 2.0212,
+      "step": 7378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.903975674261632e-05,
+      "loss": 1.5876,
+      "step": 7379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9039500362690976e-05,
+      "loss": 1.6853,
+      "step": 7380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.9039243950270766e-05,
+      "loss": 1.8388,
+      "step": 7381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9038987505356614e-05,
+      "loss": 2.1696,
+      "step": 7382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9038731027949437e-05,
+      "loss": 1.8051,
+      "step": 7383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.903847451805016e-05,
+      "loss": 2.8745,
+      "step": 7384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9038217975659705e-05,
+      "loss": 1.7283,
+      "step": 7385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.9037961400778995e-05,
+      "loss": 1.8968,
+      "step": 7386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.903770479340895e-05,
+      "loss": 1.9428,
+      "step": 7387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9037448153550495e-05,
+      "loss": 1.94,
+      "step": 7388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.903719148120455e-05,
+      "loss": 1.9139,
+      "step": 7389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.903693477637204e-05,
+      "loss": 1.9456,
+      "step": 7390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9036678039053887e-05,
+      "loss": 2.2453,
+      "step": 7391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9036421269251012e-05,
+      "loss": 2.0274,
+      "step": 7392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9036164466964338e-05,
+      "loss": 1.8856,
+      "step": 7393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.903590763219479e-05,
+      "loss": 1.8359,
+      "step": 7394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9035650764943296e-05,
+      "loss": 2.0737,
+      "step": 7395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.903539386521077e-05,
+      "loss": 1.8483,
+      "step": 7396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.903513693299814e-05,
+      "loss": 2.4179,
+      "step": 7397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9034879968306333e-05,
+      "loss": 2.075,
+      "step": 7398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9034622971136267e-05,
+      "loss": 1.8545,
+      "step": 7399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9034365941488867e-05,
+      "loss": 1.4822,
+      "step": 7400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.903410887936506e-05,
+      "loss": 1.8215,
+      "step": 7401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.9033851784765763e-05,
+      "loss": 1.8365,
+      "step": 7402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.903359465769191e-05,
+      "loss": 2.1832,
+      "step": 7403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.903333749814442e-05,
+      "loss": 1.8284,
+      "step": 7404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.9033080306124217e-05,
+      "loss": 1.7546,
+      "step": 7405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.903282308163223e-05,
+      "loss": 3.2808,
+      "step": 7406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9032565824669376e-05,
+      "loss": 2.058,
+      "step": 7407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9032308535236585e-05,
+      "loss": 2.0322,
+      "step": 7408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.903205121333478e-05,
+      "loss": 2.0061,
+      "step": 7409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9031793858964887e-05,
+      "loss": 1.7842,
+      "step": 7410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9031536472127838e-05,
+      "loss": 1.7525,
+      "step": 7411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9031279052824543e-05,
+      "loss": 1.8965,
+      "step": 7412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.903102160105594e-05,
+      "loss": 1.7044,
+      "step": 7413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9030764116822946e-05,
+      "loss": 1.9187,
+      "step": 7414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9030506600126492e-05,
+      "loss": 1.9998,
+      "step": 7415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9030249050967504e-05,
+      "loss": 1.9084,
+      "step": 7416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9029991469346907e-05,
+      "loss": 2.0245,
+      "step": 7417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9029733855265625e-05,
+      "loss": 2.0154,
+      "step": 7418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9029476208724584e-05,
+      "loss": 2.2763,
+      "step": 7419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9029218529724715e-05,
+      "loss": 1.9165,
+      "step": 7420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.9028960818266937e-05,
+      "loss": 2.1338,
+      "step": 7421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.902870307435218e-05,
+      "loss": 2.0282,
+      "step": 7422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.902844529798137e-05,
+      "loss": 1.9105,
+      "step": 7423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9028187489155438e-05,
+      "loss": 1.9645,
+      "step": 7424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.902792964787531e-05,
+      "loss": 1.8707,
+      "step": 7425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.90276717741419e-05,
+      "loss": 2.2086,
+      "step": 7426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9027413867956154e-05,
+      "loss": 1.9566,
+      "step": 7427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9027155929318986e-05,
+      "loss": 1.8817,
+      "step": 7428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.902689795823133e-05,
+      "loss": 1.8158,
+      "step": 7429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9026639954694106e-05,
+      "loss": 1.7813,
+      "step": 7430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9026381918708248e-05,
+      "loss": 2.1957,
+      "step": 7431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.902612385027468e-05,
+      "loss": 1.9045,
+      "step": 7432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9025865749394336e-05,
+      "loss": 2.0113,
+      "step": 7433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9025607616068135e-05,
+      "loss": 2.1509,
+      "step": 7434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.9025349450297012e-05,
+      "loss": 1.8774,
+      "step": 7435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.902509125208189e-05,
+      "loss": 2.0759,
+      "step": 7436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9024833021423704e-05,
+      "loss": 1.7612,
+      "step": 7437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9024574758323373e-05,
+      "loss": 1.6934,
+      "step": 7438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.902431646278183e-05,
+      "loss": 1.7852,
+      "step": 7439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.9024058134800006e-05,
+      "loss": 1.9644,
+      "step": 7440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.9023799774378825e-05,
+      "loss": 2.0945,
+      "step": 7441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.902354138151922e-05,
+      "loss": 1.7498,
+      "step": 7442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.9023282956222117e-05,
+      "loss": 2.8443,
+      "step": 7443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9023024498488444e-05,
+      "loss": 1.8855,
+      "step": 7444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.902276600831913e-05,
+      "loss": 1.806,
+      "step": 7445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.902250748571511e-05,
+      "loss": 1.8027,
+      "step": 7446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.902224893067731e-05,
+      "loss": 1.7,
+      "step": 7447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.9021990343206655e-05,
+      "loss": 1.9443,
+      "step": 7448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.9021731723304083e-05,
+      "loss": 2.1831,
+      "step": 7449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.9021473070970512e-05,
+      "loss": 1.8716,
+      "step": 7450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.9021214386206887e-05,
+      "loss": 1.7953,
+      "step": 7451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9020955669014126e-05,
+      "loss": 1.8474,
+      "step": 7452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.902069691939316e-05,
+      "loss": 1.9732,
+      "step": 7453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9020438137344924e-05,
+      "loss": 2.1431,
+      "step": 7454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.902017932287035e-05,
+      "loss": 1.8586,
+      "step": 7455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9019920475970363e-05,
+      "loss": 2.1044,
+      "step": 7456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.901966159664589e-05,
+      "loss": 1.5531,
+      "step": 7457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.9019402684897873e-05,
+      "loss": 1.99,
+      "step": 7458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.901914374072723e-05,
+      "loss": 1.5779,
+      "step": 7459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.90188847641349e-05,
+      "loss": 1.9737,
+      "step": 7460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.9018625755121815e-05,
+      "loss": 1.9207,
+      "step": 7461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9018366713688902e-05,
+      "loss": 2.0473,
+      "step": 7462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.9018107639837097e-05,
+      "loss": 1.8809,
+      "step": 7463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.901784853356732e-05,
+      "loss": 1.6158,
+      "step": 7464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9017589394880515e-05,
+      "loss": 1.8884,
+      "step": 7465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9017330223777608e-05,
+      "loss": 1.963,
+      "step": 7466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9017071020259532e-05,
+      "loss": 2.0412,
+      "step": 7467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.901681178432722e-05,
+      "loss": 1.6231,
+      "step": 7468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9016552515981598e-05,
+      "loss": 1.9216,
+      "step": 7469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9016293215223605e-05,
+      "loss": 2.0742,
+      "step": 7470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.375,
+      "learning_rate": 1.9016033882054167e-05,
+      "loss": 1.8981,
+      "step": 7471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.9015774516474224e-05,
+      "loss": 2.0892,
+      "step": 7472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.90155151184847e-05,
+      "loss": 2.2754,
+      "step": 7473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.901525568808653e-05,
+      "loss": 1.5495,
+      "step": 7474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.9014996225280653e-05,
+      "loss": 2.1371,
+      "step": 7475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.9014736730067995e-05,
+      "loss": 2.0626,
+      "step": 7476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.901447720244949e-05,
+      "loss": 1.9602,
+      "step": 7477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.901421764242607e-05,
+      "loss": 1.8485,
+      "step": 7478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.9013958049998667e-05,
+      "loss": 2.2852,
+      "step": 7479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.901369842516822e-05,
+      "loss": 1.7896,
+      "step": 7480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.901343876793566e-05,
+      "loss": 1.9645,
+      "step": 7481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9013179078301916e-05,
+      "loss": 1.7143,
+      "step": 7482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.9012919356267927e-05,
+      "loss": 2.1586,
+      "step": 7483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9012659601834625e-05,
+      "loss": 1.9085,
+      "step": 7484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.901239981500294e-05,
+      "loss": 2.1767,
+      "step": 7485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9012139995773814e-05,
+      "loss": 1.6844,
+      "step": 7486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.901188014414817e-05,
+      "loss": 1.5445,
+      "step": 7487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9011620260126953e-05,
+      "loss": 1.7889,
+      "step": 7488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.901136034371109e-05,
+      "loss": 1.6437,
+      "step": 7489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9011100394901517e-05,
+      "loss": 1.8605,
+      "step": 7490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.901084041369917e-05,
+      "loss": 1.7041,
+      "step": 7491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.9010580400104987e-05,
+      "loss": 2.9986,
+      "step": 7492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.9010320354119892e-05,
+      "loss": 2.0239,
+      "step": 7493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.9010060275744827e-05,
+      "loss": 2.1535,
+      "step": 7494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.900980016498073e-05,
+      "loss": 1.9385,
+      "step": 7495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9009540021828528e-05,
+      "loss": 1.7258,
+      "step": 7496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9009279846289165e-05,
+      "loss": 2.0638,
+      "step": 7497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9009019638363568e-05,
+      "loss": 1.5899,
+      "step": 7498
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.9008759398052676e-05,
+      "loss": 2.1821,
+      "step": 7499
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.9008499125357424e-05,
+      "loss": 1.9217,
+      "step": 7500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.900823882027875e-05,
+      "loss": 1.8332,
+      "step": 7501
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.9007978482817587e-05,
+      "loss": 1.834,
+      "step": 7502
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.9007718112974873e-05,
+      "loss": 1.8707,
+      "step": 7503
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9007457710751543e-05,
+      "loss": 1.7022,
+      "step": 7504
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.900719727614853e-05,
+      "loss": 1.9652,
+      "step": 7505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9006936809166774e-05,
+      "loss": 2.0148,
+      "step": 7506
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.900667630980721e-05,
+      "loss": 2.1365,
+      "step": 7507
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9006415778070773e-05,
+      "loss": 2.0995,
+      "step": 7508
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.9006155213958406e-05,
+      "loss": 1.5493,
+      "step": 7509
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.9005894617471037e-05,
+      "loss": 1.6437,
+      "step": 7510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.900563398860961e-05,
+      "loss": 1.4163,
+      "step": 7511
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9005373327375055e-05,
+      "loss": 1.7903,
+      "step": 7512
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9005112633768315e-05,
+      "loss": 2.1632,
+      "step": 7513
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.9004851907790323e-05,
+      "loss": 1.7037,
+      "step": 7514
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.900459114944202e-05,
+      "loss": 2.6385,
+      "step": 7515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9004330358724337e-05,
+      "loss": 1.5825,
+      "step": 7516
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.900406953563822e-05,
+      "loss": 1.7914,
+      "step": 7517
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9003808680184602e-05,
+      "loss": 1.6858,
+      "step": 7518
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.900354779236442e-05,
+      "loss": 1.6098,
+      "step": 7519
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.9003286872178614e-05,
+      "loss": 1.7221,
+      "step": 7520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9003025919628122e-05,
+      "loss": 1.9138,
+      "step": 7521
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9002764934713876e-05,
+      "loss": 1.9731,
+      "step": 7522
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9002503917436826e-05,
+      "loss": 1.8612,
+      "step": 7523
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.90022428677979e-05,
+      "loss": 2.6349,
+      "step": 7524
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.9001981785798042e-05,
+      "loss": 2.0667,
+      "step": 7525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9001720671438185e-05,
+      "loss": 1.548,
+      "step": 7526
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.900145952471927e-05,
+      "loss": 1.8661,
+      "step": 7527
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.900119834564224e-05,
+      "loss": 1.9947,
+      "step": 7528
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.900093713420803e-05,
+      "loss": 1.9385,
+      "step": 7529
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.900067589041758e-05,
+      "loss": 1.8624,
+      "step": 7530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9000414614271827e-05,
+      "loss": 1.9215,
+      "step": 7531
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.9000153305771714e-05,
+      "loss": 2.1111,
+      "step": 7532
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8999891964918178e-05,
+      "loss": 1.8045,
+      "step": 7533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8999630591712157e-05,
+      "loss": 1.8035,
+      "step": 7534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8999369186154594e-05,
+      "loss": 1.9495,
+      "step": 7535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8999107748246427e-05,
+      "loss": 1.8768,
+      "step": 7536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8998846277988598e-05,
+      "loss": 1.786,
+      "step": 7537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8998584775382038e-05,
+      "loss": 2.0238,
+      "step": 7538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.89983232404277e-05,
+      "loss": 1.9568,
+      "step": 7539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.899806167312652e-05,
+      "loss": 1.7448,
+      "step": 7540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.899780007347943e-05,
+      "loss": 2.023,
+      "step": 7541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.899753844148738e-05,
+      "loss": 1.7517,
+      "step": 7542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8997276777151302e-05,
+      "loss": 1.6232,
+      "step": 7543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.899701508047215e-05,
+      "loss": 1.7653,
+      "step": 7544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8996753351450852e-05,
+      "loss": 1.8782,
+      "step": 7545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8996491590088354e-05,
+      "loss": 2.0778,
+      "step": 7546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8996229796385597e-05,
+      "loss": 1.8432,
+      "step": 7547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.8995967970343518e-05,
+      "loss": 2.0717,
+      "step": 7548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.899570611196307e-05,
+      "loss": 1.6605,
+      "step": 7549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8995444221245177e-05,
+      "loss": 2.1506,
+      "step": 7550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.899518229819079e-05,
+      "loss": 1.9899,
+      "step": 7551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8994920342800856e-05,
+      "loss": 1.6613,
+      "step": 7552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8994658355076304e-05,
+      "loss": 2.1118,
+      "step": 7553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.899439633501809e-05,
+      "loss": 1.7819,
+      "step": 7554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8994134282627144e-05,
+      "loss": 1.8315,
+      "step": 7555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.899387219790441e-05,
+      "loss": 2.2124,
+      "step": 7556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8993610080850838e-05,
+      "loss": 2.1699,
+      "step": 7557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8993347931467358e-05,
+      "loss": 1.9202,
+      "step": 7558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.899308574975492e-05,
+      "loss": 1.8318,
+      "step": 7559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8992823535714473e-05,
+      "loss": 1.9054,
+      "step": 7560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8992561289346944e-05,
+      "loss": 1.9865,
+      "step": 7561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8992299010653288e-05,
+      "loss": 1.5124,
+      "step": 7562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8992036699634444e-05,
+      "loss": 1.7471,
+      "step": 7563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8991774356291354e-05,
+      "loss": 1.8291,
+      "step": 7564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.899151198062496e-05,
+      "loss": 1.7403,
+      "step": 7565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.8991249572636207e-05,
+      "loss": 1.6767,
+      "step": 7566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.899098713232604e-05,
+      "loss": 1.8201,
+      "step": 7567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.89907246596954e-05,
+      "loss": 1.9751,
+      "step": 7568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8990462154745232e-05,
+      "loss": 1.9287,
+      "step": 7569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8990199617476475e-05,
+      "loss": 1.8895,
+      "step": 7570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8989937047890078e-05,
+      "loss": 2.0759,
+      "step": 7571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8989674445986988e-05,
+      "loss": 1.7194,
+      "step": 7572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8989411811768138e-05,
+      "loss": 1.7879,
+      "step": 7573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8989149145234483e-05,
+      "loss": 1.5872,
+      "step": 7574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8988886446386958e-05,
+      "loss": 1.9491,
+      "step": 7575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.898862371522652e-05,
+      "loss": 1.502,
+      "step": 7576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8988360951754097e-05,
+      "loss": 1.6398,
+      "step": 7577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8988098155970644e-05,
+      "loss": 1.9596,
+      "step": 7578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8987835327877106e-05,
+      "loss": 2.06,
+      "step": 7579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8987572467474425e-05,
+      "loss": 1.872,
+      "step": 7580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8987309574763546e-05,
+      "loss": 1.8886,
+      "step": 7581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8987046649745417e-05,
+      "loss": 2.0587,
+      "step": 7582
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8986783692420977e-05,
+      "loss": 1.824,
+      "step": 7583
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8986520702791175e-05,
+      "loss": 1.6762,
+      "step": 7584
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8986257680856956e-05,
+      "loss": 2.0701,
+      "step": 7585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.898599462661927e-05,
+      "loss": 1.9397,
+      "step": 7586
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8985731540079054e-05,
+      "loss": 1.8515,
+      "step": 7587
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.898546842123726e-05,
+      "loss": 1.6618,
+      "step": 7588
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8985205270094832e-05,
+      "loss": 2.3726,
+      "step": 7589
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8984942086652717e-05,
+      "loss": 2.3369,
+      "step": 7590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.898467887091186e-05,
+      "loss": 1.7295,
+      "step": 7591
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8984415622873208e-05,
+      "loss": 1.9037,
+      "step": 7592
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8984152342537704e-05,
+      "loss": 1.9449,
+      "step": 7593
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8983889029906297e-05,
+      "loss": 1.8339,
+      "step": 7594
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8983625684979933e-05,
+      "loss": 1.8629,
+      "step": 7595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8983362307759563e-05,
+      "loss": 1.7071,
+      "step": 7596
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8983098898246127e-05,
+      "loss": 1.8549,
+      "step": 7597
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.898283545644058e-05,
+      "loss": 2.009,
+      "step": 7598
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8982571982343856e-05,
+      "loss": 1.7897,
+      "step": 7599
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8982308475956915e-05,
+      "loss": 1.5722,
+      "step": 7600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.89820449372807e-05,
+      "loss": 2.1162,
+      "step": 7601
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8981781366316152e-05,
+      "loss": 2.0188,
+      "step": 7602
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8981517763064228e-05,
+      "loss": 2.0977,
+      "step": 7603
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8981254127525874e-05,
+      "loss": 1.8881,
+      "step": 7604
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8980990459702032e-05,
+      "loss": 1.8909,
+      "step": 7605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8980726759593655e-05,
+      "loss": 1.7282,
+      "step": 7606
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8980463027201685e-05,
+      "loss": 2.0689,
+      "step": 7607
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.898019926252708e-05,
+      "loss": 1.8205,
+      "step": 7608
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.897993546557078e-05,
+      "loss": 2.0686,
+      "step": 7609
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8979671636333737e-05,
+      "loss": 1.8641,
+      "step": 7610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8979407774816896e-05,
+      "loss": 2.0982,
+      "step": 7611
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.897914388102121e-05,
+      "loss": 2.0664,
+      "step": 7612
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8978879954947623e-05,
+      "loss": 1.6654,
+      "step": 7613
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8978615996597084e-05,
+      "loss": 2.1127,
+      "step": 7614
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.897835200597055e-05,
+      "loss": 1.7714,
+      "step": 7615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8978087983068958e-05,
+      "loss": 1.7026,
+      "step": 7616
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8977823927893263e-05,
+      "loss": 1.8664,
+      "step": 7617
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8977559840444417e-05,
+      "loss": 2.1557,
+      "step": 7618
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.897729572072337e-05,
+      "loss": 1.8127,
+      "step": 7619
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8977031568731063e-05,
+      "loss": 2.0902,
+      "step": 7620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.8976767384468452e-05,
+      "loss": 2.0082,
+      "step": 7621
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8976503167936484e-05,
+      "loss": 2.1068,
+      "step": 7622
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.897623891913611e-05,
+      "loss": 1.9107,
+      "step": 7623
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.897597463806828e-05,
+      "loss": 1.7952,
+      "step": 7624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8975710324733946e-05,
+      "loss": 1.6317,
+      "step": 7625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8975445979134055e-05,
+      "loss": 1.9491,
+      "step": 7626
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8975181601269555e-05,
+      "loss": 2.1104,
+      "step": 7627
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8974917191141404e-05,
+      "loss": 1.86,
+      "step": 7628
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8974652748750547e-05,
+      "loss": 1.9517,
+      "step": 7629
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8974388274097934e-05,
+      "loss": 1.9044,
+      "step": 7630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.897412376718452e-05,
+      "loss": 1.9379,
+      "step": 7631
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8973859228011253e-05,
+      "loss": 2.1419,
+      "step": 7632
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8973594656579084e-05,
+      "loss": 1.946,
+      "step": 7633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8973330052888966e-05,
+      "loss": 1.7362,
+      "step": 7634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.897306541694185e-05,
+      "loss": 1.963,
+      "step": 7635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.897280074873868e-05,
+      "loss": 1.7514,
+      "step": 7636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8972536048280414e-05,
+      "loss": 1.7635,
+      "step": 7637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8972271315568008e-05,
+      "loss": 2.2206,
+      "step": 7638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.89720065506024e-05,
+      "loss": 1.9823,
+      "step": 7639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.897174175338456e-05,
+      "loss": 1.5275,
+      "step": 7640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8971476923915425e-05,
+      "loss": 1.9103,
+      "step": 7641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.897121206219595e-05,
+      "loss": 1.9423,
+      "step": 7642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8970947168227093e-05,
+      "loss": 1.9339,
+      "step": 7643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8970682242009807e-05,
+      "loss": 2.0662,
+      "step": 7644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.897041728354503e-05,
+      "loss": 2.145,
+      "step": 7645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.897015229283373e-05,
+      "loss": 1.6925,
+      "step": 7646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8969887269876854e-05,
+      "loss": 1.7374,
+      "step": 7647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.8969622214675353e-05,
+      "loss": 2.0577,
+      "step": 7648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.896935712723018e-05,
+      "loss": 1.9169,
+      "step": 7649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8969092007542292e-05,
+      "loss": 1.9978,
+      "step": 7650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.896882685561264e-05,
+      "loss": 1.7936,
+      "step": 7651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8968561671442174e-05,
+      "loss": 1.7897,
+      "step": 7652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.896829645503185e-05,
+      "loss": 1.9181,
+      "step": 7653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8968031206382624e-05,
+      "loss": 1.8909,
+      "step": 7654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8967765925495445e-05,
+      "loss": 1.8647,
+      "step": 7655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.896750061237127e-05,
+      "loss": 1.6264,
+      "step": 7656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.896723526701105e-05,
+      "loss": 2.0572,
+      "step": 7657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8966969889415737e-05,
+      "loss": 2.0957,
+      "step": 7658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8966704479586287e-05,
+      "loss": 2.0574,
+      "step": 7659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.896643903752366e-05,
+      "loss": 1.9356,
+      "step": 7660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8966173563228803e-05,
+      "loss": 1.8144,
+      "step": 7661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8965908056702672e-05,
+      "loss": 1.817,
+      "step": 7662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8965642517946222e-05,
+      "loss": 2.1894,
+      "step": 7663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8965376946960406e-05,
+      "loss": 1.7335,
+      "step": 7664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8965111343746186e-05,
+      "loss": 1.8579,
+      "step": 7665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8964845708304507e-05,
+      "loss": 1.6404,
+      "step": 7666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8964580040636328e-05,
+      "loss": 1.7765,
+      "step": 7667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8964314340742604e-05,
+      "loss": 2.1395,
+      "step": 7668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8964048608624287e-05,
+      "loss": 1.8939,
+      "step": 7669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.896378284428234e-05,
+      "loss": 2.0179,
+      "step": 7670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8963517047717708e-05,
+      "loss": 2.1035,
+      "step": 7671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8963251218931357e-05,
+      "loss": 1.9262,
+      "step": 7672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8962985357924235e-05,
+      "loss": 1.9996,
+      "step": 7673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.89627194646973e-05,
+      "loss": 2.0832,
+      "step": 7674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.896245353925151e-05,
+      "loss": 2.088,
+      "step": 7675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8962187581587818e-05,
+      "loss": 1.7884,
+      "step": 7676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.896192159170718e-05,
+      "loss": 1.5488,
+      "step": 7677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8961655569610557e-05,
+      "loss": 1.8643,
+      "step": 7678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.89613895152989e-05,
+      "loss": 1.9277,
+      "step": 7679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8961123428773167e-05,
+      "loss": 1.928,
+      "step": 7680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8960857310034312e-05,
+      "loss": 2.0679,
+      "step": 7681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.89605911590833e-05,
+      "loss": 1.734,
+      "step": 7682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8960324975921076e-05,
+      "loss": 1.9311,
+      "step": 7683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.8960058760548607e-05,
+      "loss": 1.7146,
+      "step": 7684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.895979251296684e-05,
+      "loss": 1.696,
+      "step": 7685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8959526233176744e-05,
+      "loss": 1.695,
+      "step": 7686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8959259921179268e-05,
+      "loss": 2.1245,
+      "step": 7687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.895899357697537e-05,
+      "loss": 2.456,
+      "step": 7688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.895872720056601e-05,
+      "loss": 1.6293,
+      "step": 7689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.8958460791952142e-05,
+      "loss": 1.6521,
+      "step": 7690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8958194351134728e-05,
+      "loss": 1.8451,
+      "step": 7691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8957927878114724e-05,
+      "loss": 1.7441,
+      "step": 7692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8957661372893092e-05,
+      "loss": 1.8619,
+      "step": 7693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.895739483547078e-05,
+      "loss": 1.9105,
+      "step": 7694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8957128265848753e-05,
+      "loss": 1.9584,
+      "step": 7695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8956861664027967e-05,
+      "loss": 1.682,
+      "step": 7696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8956595030009385e-05,
+      "loss": 1.8598,
+      "step": 7697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.895632836379396e-05,
+      "loss": 1.9153,
+      "step": 7698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8956061665382652e-05,
+      "loss": 1.7759,
+      "step": 7699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8955794934776422e-05,
+      "loss": 1.8765,
+      "step": 7700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8955528171976225e-05,
+      "loss": 1.9471,
+      "step": 7701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8955261376983026e-05,
+      "loss": 2.0134,
+      "step": 7702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8954994549797777e-05,
+      "loss": 1.9575,
+      "step": 7703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.895472769042144e-05,
+      "loss": 2.0406,
+      "step": 7704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8954460798854977e-05,
+      "loss": 1.8674,
+      "step": 7705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8954193875099346e-05,
+      "loss": 2.0097,
+      "step": 7706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.89539269191555e-05,
+      "loss": 1.8231,
+      "step": 7707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8953659931024408e-05,
+      "loss": 2.0239,
+      "step": 7708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8953392910707028e-05,
+      "loss": 2.2214,
+      "step": 7709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8953125858204314e-05,
+      "loss": 1.6169,
+      "step": 7710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8952858773517234e-05,
+      "loss": 1.994,
+      "step": 7711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8952591656646738e-05,
+      "loss": 1.917,
+      "step": 7712
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.8952324507593798e-05,
+      "loss": 2.1314,
+      "step": 7713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.895205732635937e-05,
+      "loss": 1.718,
+      "step": 7714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8951790112944406e-05,
+      "loss": 2.246,
+      "step": 7715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.895152286734988e-05,
+      "loss": 1.9817,
+      "step": 7716
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.895125558957674e-05,
+      "loss": 1.8917,
+      "step": 7717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8950988279625958e-05,
+      "loss": 2.1274,
+      "step": 7718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8950720937498488e-05,
+      "loss": 1.9728,
+      "step": 7719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8950453563195292e-05,
+      "loss": 1.8355,
+      "step": 7720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8950186156717337e-05,
+      "loss": 1.992,
+      "step": 7721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8949918718065573e-05,
+      "loss": 1.8143,
+      "step": 7722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.894965124724097e-05,
+      "loss": 1.909,
+      "step": 7723
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8949383744244488e-05,
+      "loss": 1.8165,
+      "step": 7724
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.894911620907709e-05,
+      "loss": 1.8999,
+      "step": 7725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8948848641739732e-05,
+      "loss": 1.8549,
+      "step": 7726
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8948581042233383e-05,
+      "loss": 1.7009,
+      "step": 7727
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8948313410559e-05,
+      "loss": 1.8069,
+      "step": 7728
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.8948045746717546e-05,
+      "loss": 1.5245,
+      "step": 7729
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8947778050709984e-05,
+      "loss": 1.8018,
+      "step": 7730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8947510322537273e-05,
+      "loss": 1.8855,
+      "step": 7731
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.8947242562200384e-05,
+      "loss": 2.0232,
+      "step": 7732
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.894697476970027e-05,
+      "loss": 1.9586,
+      "step": 7733
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8946706945037897e-05,
+      "loss": 1.82,
+      "step": 7734
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.894643908821423e-05,
+      "loss": 1.9751,
+      "step": 7735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8946171199230232e-05,
+      "loss": 1.9043,
+      "step": 7736
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8945903278086862e-05,
+      "loss": 1.7108,
+      "step": 7737
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8945635324785086e-05,
+      "loss": 1.6786,
+      "step": 7738
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8945367339325867e-05,
+      "loss": 1.9682,
+      "step": 7739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8945099321710167e-05,
+      "loss": 1.8531,
+      "step": 7740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8944831271938952e-05,
+      "loss": 1.7987,
+      "step": 7741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.8944563190013184e-05,
+      "loss": 2.2112,
+      "step": 7742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8944295075933824e-05,
+      "loss": 1.8871,
+      "step": 7743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.894402692970184e-05,
+      "loss": 2.0406,
+      "step": 7744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8943758751318195e-05,
+      "loss": 1.7436,
+      "step": 7745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.894349054078385e-05,
+      "loss": 1.942,
+      "step": 7746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8943222298099773e-05,
+      "loss": 1.7867,
+      "step": 7747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.894295402326693e-05,
+      "loss": 1.6676,
+      "step": 7748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.894268571628628e-05,
+      "loss": 1.894,
+      "step": 7749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8942417377158787e-05,
+      "loss": 2.1583,
+      "step": 7750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8942149005885422e-05,
+      "loss": 1.5972,
+      "step": 7751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8941880602467143e-05,
+      "loss": 2.1633,
+      "step": 7752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.894161216690492e-05,
+      "loss": 1.8954,
+      "step": 7753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8941343699199716e-05,
+      "loss": 1.9068,
+      "step": 7754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8941075199352496e-05,
+      "loss": 2.3421,
+      "step": 7755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8940806667364225e-05,
+      "loss": 1.9775,
+      "step": 7756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8940538103235867e-05,
+      "loss": 1.9845,
+      "step": 7757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.894026950696839e-05,
+      "loss": 2.2682,
+      "step": 7758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8940000878562758e-05,
+      "loss": 1.9445,
+      "step": 7759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8939732218019936e-05,
+      "loss": 1.9444,
+      "step": 7760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8939463525340892e-05,
+      "loss": 1.9361,
+      "step": 7761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8939194800526593e-05,
+      "loss": 2.1324,
+      "step": 7762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8938926043578002e-05,
+      "loss": 1.9655,
+      "step": 7763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8938657254496085e-05,
+      "loss": 2.0494,
+      "step": 7764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.893838843328181e-05,
+      "loss": 1.7867,
+      "step": 7765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.893811957993614e-05,
+      "loss": 2.0664,
+      "step": 7766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8937850694460044e-05,
+      "loss": 1.7988,
+      "step": 7767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.893758177685449e-05,
+      "loss": 1.9432,
+      "step": 7768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8937312827120445e-05,
+      "loss": 1.6078,
+      "step": 7769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.893704384525887e-05,
+      "loss": 2.2523,
+      "step": 7770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.893677483127074e-05,
+      "loss": 1.6924,
+      "step": 7771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8936505785157012e-05,
+      "loss": 1.7777,
+      "step": 7772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8936236706918663e-05,
+      "loss": 1.8705,
+      "step": 7773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8935967596556658e-05,
+      "loss": 1.6209,
+      "step": 7774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.893569845407196e-05,
+      "loss": 1.8837,
+      "step": 7775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.893542927946554e-05,
+      "loss": 1.7055,
+      "step": 7776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8935160072738367e-05,
+      "loss": 1.8827,
+      "step": 7777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8934890833891404e-05,
+      "loss": 1.999,
+      "step": 7778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8934621562925622e-05,
+      "loss": 1.637,
+      "step": 7779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.893435225984199e-05,
+      "loss": 2.0357,
+      "step": 7780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8934082924641473e-05,
+      "loss": 2.0909,
+      "step": 7781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8933813557325044e-05,
+      "loss": 1.8008,
+      "step": 7782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.893354415789366e-05,
+      "loss": 1.7024,
+      "step": 7783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.893327472634831e-05,
+      "loss": 1.8754,
+      "step": 7784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.893300526268994e-05,
+      "loss": 2.1343,
+      "step": 7785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.893273576691953e-05,
+      "loss": 1.6447,
+      "step": 7786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8932466239038047e-05,
+      "loss": 1.9465,
+      "step": 7787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8932196679046463e-05,
+      "loss": 1.8827,
+      "step": 7788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.893192708694574e-05,
+      "loss": 2.1395,
+      "step": 7789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.893165746273686e-05,
+      "loss": 1.772,
+      "step": 7790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8931387806420778e-05,
+      "loss": 1.9857,
+      "step": 7791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.893111811799847e-05,
+      "loss": 1.8247,
+      "step": 7792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.89308483974709e-05,
+      "loss": 2.0113,
+      "step": 7793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8930578644839052e-05,
+      "loss": 2.0275,
+      "step": 7794
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8930308860103878e-05,
+      "loss": 2.1029,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8930039043266358e-05,
+      "loss": 1.6237,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8929769194327462e-05,
+      "loss": 1.818,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8929499313288152e-05,
+      "loss": 1.8627,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.8929229400149407e-05,
+      "loss": 1.998,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8928959454912196e-05,
+      "loss": 1.7748,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.892868947757748e-05,
+      "loss": 1.9719,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8928419468146245e-05,
+      "loss": 1.8161,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.892814942661945e-05,
+      "loss": 1.7689,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.892787935299807e-05,
+      "loss": 1.6139,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.8927609247283073e-05,
+      "loss": 1.8633,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8927339109475435e-05,
+      "loss": 1.9847,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8927068939576124e-05,
+      "loss": 1.6937,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.892679873758611e-05,
+      "loss": 1.8697,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8926528503506368e-05,
+      "loss": 1.4287,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8926258237337863e-05,
+      "loss": 1.7928,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.8925987939081572e-05,
+      "loss": 1.5511,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8925717608738466e-05,
+      "loss": 2.05,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8925447246309516e-05,
+      "loss": 2.0709,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8925176851795696e-05,
+      "loss": 2.0463,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.892490642519797e-05,
+      "loss": 1.6562,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8924635966517323e-05,
+      "loss": 1.9356,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.892436547575472e-05,
+      "loss": 1.753,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8924094952911128e-05,
+      "loss": 2.0975,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.892382439798753e-05,
+      "loss": 2.0138,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8923553810984893e-05,
+      "loss": 1.6333,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8923283191904187e-05,
+      "loss": 2.1407,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.892301254074639e-05,
+      "loss": 2.0691,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8922741857512476e-05,
+      "loss": 1.8697,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8922471142203413e-05,
+      "loss": 2.0143,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8922200394820173e-05,
+      "loss": 1.664,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.892192961536374e-05,
+      "loss": 1.954,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8921658803835074e-05,
+      "loss": 1.9426,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8921387960235156e-05,
+      "loss": 1.9324,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8921117084564958e-05,
+      "loss": 1.8127,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8920846176825452e-05,
+      "loss": 1.7208,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.892057523701762e-05,
+      "loss": 1.884,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8920304265142424e-05,
+      "loss": 1.9363,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8920033261200846e-05,
+      "loss": 1.8579,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8919762225193857e-05,
+      "loss": 1.8791,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8919491157122433e-05,
+      "loss": 1.9954,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8919220056987548e-05,
+      "loss": 2.0228,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8918948924790176e-05,
+      "loss": 1.994,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8918677760531293e-05,
+      "loss": 1.8671,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.891840656421187e-05,
+      "loss": 1.9144,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8918135335832887e-05,
+      "loss": 1.9518,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8917864075395313e-05,
+      "loss": 1.9101,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.891759278290013e-05,
+      "loss": 2.1392,
+      "step": 7842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.891732145834831e-05,
+      "loss": 1.9978,
+      "step": 7843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8917050101740826e-05,
+      "loss": 2.0784,
+      "step": 7844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5,
+      "learning_rate": 1.8916778713078656e-05,
+      "loss": 1.7083,
+      "step": 7845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8916507292362776e-05,
+      "loss": 2.05,
+      "step": 7846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.891623583959416e-05,
+      "loss": 1.7904,
+      "step": 7847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8915964354773788e-05,
+      "loss": 1.9466,
+      "step": 7848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.891569283790263e-05,
+      "loss": 2.6979,
+      "step": 7849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8915421288981667e-05,
+      "loss": 1.8491,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.891514970801187e-05,
+      "loss": 1.6422,
+      "step": 7851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.891487809499422e-05,
+      "loss": 1.9176,
+      "step": 7852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.375,
+      "learning_rate": 1.891460644992969e-05,
+      "loss": 1.6102,
+      "step": 7853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.891433477281926e-05,
+      "loss": 1.8272,
+      "step": 7854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8914063063663906e-05,
+      "loss": 1.8676,
+      "step": 7855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.89137913224646e-05,
+      "loss": 1.8116,
+      "step": 7856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8913519549222322e-05,
+      "loss": 2.0079,
+      "step": 7857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8913247743938052e-05,
+      "loss": 1.926,
+      "step": 7858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8912975906612762e-05,
+      "loss": 1.7368,
+      "step": 7859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.891270403724743e-05,
+      "loss": 1.758,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.891243213584304e-05,
+      "loss": 1.8318,
+      "step": 7861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.891216020240056e-05,
+      "loss": 1.8399,
+      "step": 7862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.8911888236920975e-05,
+      "loss": 1.9721,
+      "step": 7863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8911616239405254e-05,
+      "loss": 2.0461,
+      "step": 7864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8911344209854388e-05,
+      "loss": 1.6572,
+      "step": 7865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.891107214826934e-05,
+      "loss": 2.1136,
+      "step": 7866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.89108000546511e-05,
+      "loss": 1.669,
+      "step": 7867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.8910527929000636e-05,
+      "loss": 2.0874,
+      "step": 7868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8910255771318936e-05,
+      "loss": 1.8842,
+      "step": 7869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.890998358160697e-05,
+      "loss": 2.0242,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8909711359865725e-05,
+      "loss": 1.9226,
+      "step": 7871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.890943910609617e-05,
+      "loss": 1.6809,
+      "step": 7872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.890916682029929e-05,
+      "loss": 2.043,
+      "step": 7873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8908894502476066e-05,
+      "loss": 2.0473,
+      "step": 7874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8908622152627467e-05,
+      "loss": 1.903,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.890834977075448e-05,
+      "loss": 1.9091,
+      "step": 7876
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8908077356858084e-05,
+      "loss": 1.7628,
+      "step": 7877
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.8907804910939255e-05,
+      "loss": 1.6559,
+      "step": 7878
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8907532432998976e-05,
+      "loss": 1.4661,
+      "step": 7879
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.8907259923038222e-05,
+      "loss": 1.791,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8906987381057976e-05,
+      "loss": 1.5806,
+      "step": 7881
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8906714807059218e-05,
+      "loss": 1.7739,
+      "step": 7882
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.890644220104293e-05,
+      "loss": 2.1158,
+      "step": 7883
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8906169563010078e-05,
+      "loss": 1.6653,
+      "step": 7884
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8905896892961663e-05,
+      "loss": 1.7447,
+      "step": 7885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.890562419089865e-05,
+      "loss": 1.7918,
+      "step": 7886
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8905351456822026e-05,
+      "loss": 1.9644,
+      "step": 7887
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.890507869073277e-05,
+      "loss": 1.8194,
+      "step": 7888
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8904805892631857e-05,
+      "loss": 1.6235,
+      "step": 7889
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.890453306252028e-05,
+      "loss": 1.6611,
+      "step": 7890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.890426020039901e-05,
+      "loss": 2.0171,
+      "step": 7891
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8903987306269026e-05,
+      "loss": 1.7626,
+      "step": 7892
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8903714380131316e-05,
+      "loss": 2.1214,
+      "step": 7893
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8903441421986862e-05,
+      "loss": 1.7153,
+      "step": 7894
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.890316843183664e-05,
+      "loss": 2.0057,
+      "step": 7895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8902895409681628e-05,
+      "loss": 1.8662,
+      "step": 7896
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8902622355522816e-05,
+      "loss": 1.7697,
+      "step": 7897
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8902349269361184e-05,
+      "loss": 1.712,
+      "step": 7898
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.890207615119771e-05,
+      "loss": 2.0074,
+      "step": 7899
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8901803001033378e-05,
+      "loss": 2.0642,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8901529818869167e-05,
+      "loss": 2.1153,
+      "step": 7901
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8901256604706065e-05,
+      "loss": 2.0187,
+      "step": 7902
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.890098335854505e-05,
+      "loss": 1.6929,
+      "step": 7903
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8900710080387107e-05,
+      "loss": 1.9216,
+      "step": 7904
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.890043677023321e-05,
+      "loss": 1.8681,
+      "step": 7905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8900163428084354e-05,
+      "loss": 1.945,
+      "step": 7906
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.889989005394151e-05,
+      "loss": 1.6182,
+      "step": 7907
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8899616647805667e-05,
+      "loss": 1.8398,
+      "step": 7908
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.889934320967781e-05,
+      "loss": 1.9744,
+      "step": 7909
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8899069739558915e-05,
+      "loss": 1.8829,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.889879623744997e-05,
+      "loss": 2.1463,
+      "step": 7911
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.889852270335196e-05,
+      "loss": 1.7072,
+      "step": 7912
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8898249137265862e-05,
+      "loss": 1.7764,
+      "step": 7913
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8897975539192662e-05,
+      "loss": 1.7411,
+      "step": 7914
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.889770190913335e-05,
+      "loss": 2.0221,
+      "step": 7915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.88974282470889e-05,
+      "loss": 1.8614,
+      "step": 7916
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8897154553060297e-05,
+      "loss": 1.8912,
+      "step": 7917
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8896880827048534e-05,
+      "loss": 2.1992,
+      "step": 7918
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8896607069054582e-05,
+      "loss": 2.2005,
+      "step": 7919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8896333279079434e-05,
+      "loss": 2.0379,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8896059457124073e-05,
+      "loss": 2.1118,
+      "step": 7921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.889578560318948e-05,
+      "loss": 1.9596,
+      "step": 7922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8895511717276643e-05,
+      "loss": 1.795,
+      "step": 7923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8895237799386544e-05,
+      "loss": 1.781,
+      "step": 7924
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8894963849520168e-05,
+      "loss": 2.043,
+      "step": 7925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8894689867678505e-05,
+      "loss": 1.9705,
+      "step": 7926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.889441585386253e-05,
+      "loss": 2.1062,
+      "step": 7927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8894141808073238e-05,
+      "loss": 1.5977,
+      "step": 7928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.889386773031161e-05,
+      "loss": 2.2302,
+      "step": 7929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8893593620578627e-05,
+      "loss": 1.9061,
+      "step": 7930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.889331947887528e-05,
+      "loss": 2.1433,
+      "step": 7931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8893045305202552e-05,
+      "loss": 2.228,
+      "step": 7932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.889277109956143e-05,
+      "loss": 2.1591,
+      "step": 7933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.88924968619529e-05,
+      "loss": 1.9538,
+      "step": 7934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8892222592377942e-05,
+      "loss": 1.8115,
+      "step": 7935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.889194829083755e-05,
+      "loss": 1.8154,
+      "step": 7936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8891673957332707e-05,
+      "loss": 1.9617,
+      "step": 7937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8891399591864398e-05,
+      "loss": 2.0398,
+      "step": 7938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8891125194433612e-05,
+      "loss": 2.1524,
+      "step": 7939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8890850765041336e-05,
+      "loss": 2.0193,
+      "step": 7940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8890576303688548e-05,
+      "loss": 2.2469,
+      "step": 7941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8890301810376245e-05,
+      "loss": 1.9511,
+      "step": 7942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8890027285105406e-05,
+      "loss": 1.772,
+      "step": 7943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8889752727877023e-05,
+      "loss": 1.8949,
+      "step": 7944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8889478138692083e-05,
+      "loss": 1.9384,
+      "step": 7945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.888920351755157e-05,
+      "loss": 1.8643,
+      "step": 7946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8888928864456474e-05,
+      "loss": 1.7826,
+      "step": 7947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.888865417940778e-05,
+      "loss": 1.8947,
+      "step": 7948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8888379462406473e-05,
+      "loss": 1.9445,
+      "step": 7949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8888104713453548e-05,
+      "loss": 1.5706,
+      "step": 7950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8887829932549988e-05,
+      "loss": 1.8094,
+      "step": 7951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8887555119696783e-05,
+      "loss": 1.8744,
+      "step": 7952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8887280274894914e-05,
+      "loss": 1.996,
+      "step": 7953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.8887005398145378e-05,
+      "loss": 2.0391,
+      "step": 7954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8886730489449158e-05,
+      "loss": 1.748,
+      "step": 7955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8886455548807246e-05,
+      "loss": 2.0125,
+      "step": 7956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8886180576220625e-05,
+      "loss": 1.9198,
+      "step": 7957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8885905571690287e-05,
+      "loss": 1.9165,
+      "step": 7958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.888563053521722e-05,
+      "loss": 1.9781,
+      "step": 7959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8885355466802413e-05,
+      "loss": 2.0765,
+      "step": 7960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.888508036644685e-05,
+      "loss": 1.7891,
+      "step": 7961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.888480523415153e-05,
+      "loss": 1.7161,
+      "step": 7962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8884530069917434e-05,
+      "loss": 1.9331,
+      "step": 7963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8884254873745557e-05,
+      "loss": 1.9021,
+      "step": 7964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.888397964563688e-05,
+      "loss": 1.9109,
+      "step": 7965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.88837043855924e-05,
+      "loss": 1.716,
+      "step": 7966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.8883429093613104e-05,
+      "loss": 1.9597,
+      "step": 7967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.888315376969998e-05,
+      "loss": 1.9507,
+      "step": 7968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.888287841385402e-05,
+      "loss": 1.7241,
+      "step": 7969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.888260302607621e-05,
+      "loss": 1.8438,
+      "step": 7970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8882327606367544e-05,
+      "loss": 1.9759,
+      "step": 7971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8882052154729015e-05,
+      "loss": 1.7719,
+      "step": 7972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8881776671161606e-05,
+      "loss": 1.7086,
+      "step": 7973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.888150115566631e-05,
+      "loss": 1.9031,
+      "step": 7974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8881225608244117e-05,
+      "loss": 1.7324,
+      "step": 7975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8880950028896018e-05,
+      "loss": 1.7787,
+      "step": 7976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8880674417623007e-05,
+      "loss": 1.6393,
+      "step": 7977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.888039877442607e-05,
+      "loss": 1.8304,
+      "step": 7978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.88801230993062e-05,
+      "loss": 1.8764,
+      "step": 7979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8879847392264387e-05,
+      "loss": 2.1338,
+      "step": 7980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8879571653301623e-05,
+      "loss": 1.934,
+      "step": 7981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8879295882418898e-05,
+      "loss": 1.5249,
+      "step": 7982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8879020079617202e-05,
+      "loss": 2.1024,
+      "step": 7983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8878744244897533e-05,
+      "loss": 1.8511,
+      "step": 7984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8878468378260873e-05,
+      "loss": 2.0402,
+      "step": 7985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8878192479708226e-05,
+      "loss": 1.6509,
+      "step": 7986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8877916549240573e-05,
+      "loss": 1.9231,
+      "step": 7987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.887764058685891e-05,
+      "loss": 1.9132,
+      "step": 7988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8877364592564227e-05,
+      "loss": 1.812,
+      "step": 7989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.887708856635752e-05,
+      "loss": 1.7355,
+      "step": 7990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.375,
+      "learning_rate": 1.887681250823978e-05,
+      "loss": 2.0011,
+      "step": 7991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8876536418211994e-05,
+      "loss": 1.6881,
+      "step": 7992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8876260296275162e-05,
+      "loss": 1.7487,
+      "step": 7993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8875984142430274e-05,
+      "loss": 1.7999,
+      "step": 7994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.887570795667832e-05,
+      "loss": 1.9151,
+      "step": 7995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8875431739020292e-05,
+      "loss": 2.1663,
+      "step": 7996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.887515548945719e-05,
+      "loss": 1.9917,
+      "step": 7997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8874879207990004e-05,
+      "loss": 1.8686,
+      "step": 7998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8874602894619725e-05,
+      "loss": 1.8143,
+      "step": 7999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8874326549347346e-05,
+      "loss": 1.9073,
+      "step": 8000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8874050172173862e-05,
+      "loss": 1.7805,
+      "step": 8001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8873773763100268e-05,
+      "loss": 1.734,
+      "step": 8002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8873497322127555e-05,
+      "loss": 2.4333,
+      "step": 8003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8873220849256716e-05,
+      "loss": 1.8091,
+      "step": 8004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8872944344488747e-05,
+      "loss": 2.0188,
+      "step": 8005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8872667807824645e-05,
+      "loss": 1.6928,
+      "step": 8006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8872391239265396e-05,
+      "loss": 1.7121,
+      "step": 8007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8872114638812e-05,
+      "loss": 2.0299,
+      "step": 8008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8871838006465452e-05,
+      "loss": 2.06,
+      "step": 8009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8871561342226746e-05,
+      "loss": 1.8268,
+      "step": 8010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.887128464609687e-05,
+      "loss": 1.8939,
+      "step": 8011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8871007918076828e-05,
+      "loss": 1.8871,
+      "step": 8012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8870731158167608e-05,
+      "loss": 1.8004,
+      "step": 8013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8870454366370206e-05,
+      "loss": 1.8414,
+      "step": 8014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8870177542685622e-05,
+      "loss": 1.7626,
+      "step": 8015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8869900687114845e-05,
+      "loss": 1.8873,
+      "step": 8016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8869623799658875e-05,
+      "loss": 1.6526,
+      "step": 8017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.88693468803187e-05,
+      "loss": 1.8001,
+      "step": 8018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8869069929095326e-05,
+      "loss": 2.2429,
+      "step": 8019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.886879294598974e-05,
+      "loss": 1.9671,
+      "step": 8020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8868515931002944e-05,
+      "loss": 1.4488,
+      "step": 8021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.886823888413593e-05,
+      "loss": 2.0807,
+      "step": 8022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.886796180538969e-05,
+      "loss": 2.2341,
+      "step": 8023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8867684694765228e-05,
+      "loss": 1.7012,
+      "step": 8024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8867407552263538e-05,
+      "loss": 1.9995,
+      "step": 8025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8867130377885612e-05,
+      "loss": 1.8816,
+      "step": 8026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.886685317163245e-05,
+      "loss": 1.9222,
+      "step": 8027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8866575933505048e-05,
+      "loss": 1.6476,
+      "step": 8028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.88662986635044e-05,
+      "loss": 1.6594,
+      "step": 8029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.886602136163151e-05,
+      "loss": 1.9302,
+      "step": 8030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8865744027887365e-05,
+      "loss": 1.9354,
+      "step": 8031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8865466662272972e-05,
+      "loss": 1.8153,
+      "step": 8032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.886518926478932e-05,
+      "loss": 1.4482,
+      "step": 8033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8864911835437408e-05,
+      "loss": 1.9453,
+      "step": 8034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8864634374218237e-05,
+      "loss": 1.7733,
+      "step": 8035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.88643568811328e-05,
+      "loss": 1.8626,
+      "step": 8036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8864079356182098e-05,
+      "loss": 1.7239,
+      "step": 8037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8863801799367127e-05,
+      "loss": 1.7286,
+      "step": 8038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8863524210688884e-05,
+      "loss": 2.1105,
+      "step": 8039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8863246590148368e-05,
+      "loss": 1.7,
+      "step": 8040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.886296893774658e-05,
+      "loss": 1.807,
+      "step": 8041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8862691253484506e-05,
+      "loss": 1.9168,
+      "step": 8042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.886241353736316e-05,
+      "loss": 1.4928,
+      "step": 8043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8862135789383532e-05,
+      "loss": 1.9958,
+      "step": 8044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.886185800954662e-05,
+      "loss": 1.7582,
+      "step": 8045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8861580197853423e-05,
+      "loss": 1.9495,
+      "step": 8046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8861302354304945e-05,
+      "loss": 1.7686,
+      "step": 8047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8861024478902176e-05,
+      "loss": 1.7212,
+      "step": 8048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8860746571646124e-05,
+      "loss": 1.526,
+      "step": 8049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8860468632537784e-05,
+      "loss": 1.8361,
+      "step": 8050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8860190661578152e-05,
+      "loss": 1.6378,
+      "step": 8051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8859912658768228e-05,
+      "loss": 1.8518,
+      "step": 8052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8859634624109015e-05,
+      "loss": 1.9515,
+      "step": 8053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8859356557601512e-05,
+      "loss": 1.8706,
+      "step": 8054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.885907845924672e-05,
+      "loss": 1.8589,
+      "step": 8055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.885880032904563e-05,
+      "loss": 1.6029,
+      "step": 8056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.885852216699925e-05,
+      "loss": 1.8558,
+      "step": 8057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.885824397310858e-05,
+      "loss": 1.8143,
+      "step": 8058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8857965747374617e-05,
+      "loss": 2.3021,
+      "step": 8059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8857687489798362e-05,
+      "loss": 1.838,
+      "step": 8060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.885740920038081e-05,
+      "loss": 1.7856,
+      "step": 8061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8857130879122975e-05,
+      "loss": 1.7212,
+      "step": 8062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8856852526025845e-05,
+      "loss": 1.8496,
+      "step": 8063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8856574141090424e-05,
+      "loss": 1.7605,
+      "step": 8064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8856295724317718e-05,
+      "loss": 1.6227,
+      "step": 8065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8856017275708716e-05,
+      "loss": 2.0015,
+      "step": 8066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8855738795264432e-05,
+      "loss": 1.827,
+      "step": 8067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.885546028298586e-05,
+      "loss": 1.6639,
+      "step": 8068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8855181738874e-05,
+      "loss": 1.9262,
+      "step": 8069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.885490316292986e-05,
+      "loss": 2.3881,
+      "step": 8070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8854624555154436e-05,
+      "loss": 1.749,
+      "step": 8071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8854345915548727e-05,
+      "loss": 2.3208,
+      "step": 8072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8854067244113743e-05,
+      "loss": 1.7812,
+      "step": 8073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.885378854085048e-05,
+      "loss": 2.254,
+      "step": 8074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8853509805759942e-05,
+      "loss": 1.7012,
+      "step": 8075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8853231038843128e-05,
+      "loss": 1.7999,
+      "step": 8076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8852952240101043e-05,
+      "loss": 2.1527,
+      "step": 8077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8852673409534687e-05,
+      "loss": 2.0302,
+      "step": 8078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8852394547145065e-05,
+      "loss": 2.2592,
+      "step": 8079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 54.75,
+      "learning_rate": 1.885211565293318e-05,
+      "loss": 3.9915,
+      "step": 8080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.885183672690003e-05,
+      "loss": 1.7879,
+      "step": 8081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.885155776904662e-05,
+      "loss": 1.8491,
+      "step": 8082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8851278779373954e-05,
+      "loss": 1.7147,
+      "step": 8083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8850999757883035e-05,
+      "loss": 1.8512,
+      "step": 8084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8850720704574864e-05,
+      "loss": 2.034,
+      "step": 8085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8850441619450448e-05,
+      "loss": 1.9483,
+      "step": 8086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8850162502510784e-05,
+      "loss": 2.0499,
+      "step": 8087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.884988335375688e-05,
+      "loss": 2.1368,
+      "step": 8088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8849604173189738e-05,
+      "loss": 1.7633,
+      "step": 8089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.884932496081036e-05,
+      "loss": 2.1205,
+      "step": 8090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.8849045716619755e-05,
+      "loss": 1.8251,
+      "step": 8091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.884876644061892e-05,
+      "loss": 1.8867,
+      "step": 8092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8848487132808866e-05,
+      "loss": 1.8951,
+      "step": 8093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8848207793190593e-05,
+      "loss": 1.7782,
+      "step": 8094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8847928421765104e-05,
+      "loss": 1.7225,
+      "step": 8095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8847649018533406e-05,
+      "loss": 2.0661,
+      "step": 8096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8847369583496498e-05,
+      "loss": 1.8786,
+      "step": 8097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8847090116655394e-05,
+      "loss": 2.065,
+      "step": 8098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.884681061801109e-05,
+      "loss": 1.9233,
+      "step": 8099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8846531087564597e-05,
+      "loss": 1.88,
+      "step": 8100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8846251525316914e-05,
+      "loss": 1.8056,
+      "step": 8101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.884597193126905e-05,
+      "loss": 1.9609,
+      "step": 8102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.8845692305422008e-05,
+      "loss": 2.0183,
+      "step": 8103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8845412647776795e-05,
+      "loss": 1.7104,
+      "step": 8104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8845132958334416e-05,
+      "loss": 1.6451,
+      "step": 8105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8844853237095873e-05,
+      "loss": 1.8979,
+      "step": 8106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8844573484062177e-05,
+      "loss": 2.137,
+      "step": 8107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8844293699234334e-05,
+      "loss": 1.7255,
+      "step": 8108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8844013882613342e-05,
+      "loss": 2.0197,
+      "step": 8109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8843734034200212e-05,
+      "loss": 2.1805,
+      "step": 8110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.884345415399595e-05,
+      "loss": 1.8292,
+      "step": 8111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8843174242001563e-05,
+      "loss": 1.8242,
+      "step": 8112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8842894298218057e-05,
+      "loss": 1.8566,
+      "step": 8113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8842614322646433e-05,
+      "loss": 1.7005,
+      "step": 8114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8842334315287707e-05,
+      "loss": 1.9716,
+      "step": 8115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.884205427614288e-05,
+      "loss": 1.7847,
+      "step": 8116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8841774205212954e-05,
+      "loss": 1.7357,
+      "step": 8117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.884149410249894e-05,
+      "loss": 1.9753,
+      "step": 8118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8841213968001853e-05,
+      "loss": 1.9774,
+      "step": 8119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8840933801722686e-05,
+      "loss": 1.9746,
+      "step": 8120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8840653603662457e-05,
+      "loss": 1.9394,
+      "step": 8121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8840373373822168e-05,
+      "loss": 1.8083,
+      "step": 8122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8840093112202826e-05,
+      "loss": 1.7677,
+      "step": 8123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.883981281880544e-05,
+      "loss": 2.0407,
+      "step": 8124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8839532493631017e-05,
+      "loss": 1.9293,
+      "step": 8125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8839252136680567e-05,
+      "loss": 1.6985,
+      "step": 8126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8838971747955098e-05,
+      "loss": 1.671,
+      "step": 8127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.883869132745561e-05,
+      "loss": 1.7815,
+      "step": 8128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.883841087518312e-05,
+      "loss": 1.74,
+      "step": 8129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.883813039113864e-05,
+      "loss": 1.845,
+      "step": 8130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.883784987532316e-05,
+      "loss": 1.8173,
+      "step": 8131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8837569327737707e-05,
+      "loss": 2.1584,
+      "step": 8132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.883728874838328e-05,
+      "loss": 1.8779,
+      "step": 8133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.883700813726089e-05,
+      "loss": 1.6883,
+      "step": 8134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8836727494371545e-05,
+      "loss": 1.8808,
+      "step": 8135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8836446819716254e-05,
+      "loss": 2.0925,
+      "step": 8136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.8836166113296028e-05,
+      "loss": 2.2378,
+      "step": 8137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8835885375111874e-05,
+      "loss": 1.5185,
+      "step": 8138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.88356046051648e-05,
+      "loss": 1.9216,
+      "step": 8139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.883532380345582e-05,
+      "loss": 1.9164,
+      "step": 8140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.883504296998594e-05,
+      "loss": 2.2738,
+      "step": 8141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8834762104756167e-05,
+      "loss": 1.8179,
+      "step": 8142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8834481207767518e-05,
+      "loss": 1.6432,
+      "step": 8143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8834200279020994e-05,
+      "loss": 2.0417,
+      "step": 8144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8833919318517614e-05,
+      "loss": 1.678,
+      "step": 8145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8833638326258378e-05,
+      "loss": 2.0627,
+      "step": 8146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8833357302244304e-05,
+      "loss": 2.0366,
+      "step": 8147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8833076246476398e-05,
+      "loss": 1.8891,
+      "step": 8148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8832795158955673e-05,
+      "loss": 1.7087,
+      "step": 8149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.883251403968314e-05,
+      "loss": 1.9466,
+      "step": 8150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.883223288865981e-05,
+      "loss": 2.2427,
+      "step": 8151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8831951705886685e-05,
+      "loss": 1.8139,
+      "step": 8152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8831670491364784e-05,
+      "loss": 1.753,
+      "step": 8153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8831389245095116e-05,
+      "loss": 2.1145,
+      "step": 8154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8831107967078694e-05,
+      "loss": 1.9412,
+      "step": 8155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8830826657316528e-05,
+      "loss": 2.0314,
+      "step": 8156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8830545315809626e-05,
+      "loss": 2.0613,
+      "step": 8157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8830263942559006e-05,
+      "loss": 1.8856,
+      "step": 8158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.882998253756567e-05,
+      "loss": 1.8515,
+      "step": 8159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.882970110083064e-05,
+      "loss": 2.1981,
+      "step": 8160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8829419632354922e-05,
+      "loss": 1.8283,
+      "step": 8161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8829138132139525e-05,
+      "loss": 1.8836,
+      "step": 8162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8828856600185468e-05,
+      "loss": 1.7271,
+      "step": 8163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8828575036493757e-05,
+      "loss": 2.0382,
+      "step": 8164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.882829344106541e-05,
+      "loss": 1.9377,
+      "step": 8165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8828011813901433e-05,
+      "loss": 2.2018,
+      "step": 8166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.882773015500284e-05,
+      "loss": 1.7828,
+      "step": 8167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8827448464370648e-05,
+      "loss": 1.9476,
+      "step": 8168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8827166742005865e-05,
+      "loss": 1.7839,
+      "step": 8169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8826884987909502e-05,
+      "loss": 1.7683,
+      "step": 8170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.882660320208258e-05,
+      "loss": 2.2941,
+      "step": 8171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8826321384526104e-05,
+      "loss": 1.8583,
+      "step": 8172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.882603953524109e-05,
+      "loss": 1.8481,
+      "step": 8173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8825757654228554e-05,
+      "loss": 1.9292,
+      "step": 8174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8825475741489504e-05,
+      "loss": 2.0681,
+      "step": 8175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.882519379702496e-05,
+      "loss": 1.9584,
+      "step": 8176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8824911820835926e-05,
+      "loss": 1.8456,
+      "step": 8177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8824629812923424e-05,
+      "loss": 1.9442,
+      "step": 8178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8824347773288464e-05,
+      "loss": 1.5216,
+      "step": 8179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8824065701932062e-05,
+      "loss": 1.7655,
+      "step": 8180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.882378359885523e-05,
+      "loss": 1.6183,
+      "step": 8181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.882350146405898e-05,
+      "loss": 1.7826,
+      "step": 8182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.375,
+      "learning_rate": 1.882321929754433e-05,
+      "loss": 1.9158,
+      "step": 8183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8822937099312296e-05,
+      "loss": 1.6886,
+      "step": 8184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8822654869363886e-05,
+      "loss": 1.7961,
+      "step": 8185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.882237260770012e-05,
+      "loss": 1.9526,
+      "step": 8186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.882209031432201e-05,
+      "loss": 1.9632,
+      "step": 8187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.8821807989230574e-05,
+      "loss": 1.8253,
+      "step": 8188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8821525632426824e-05,
+      "loss": 2.1102,
+      "step": 8189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8821243243911773e-05,
+      "loss": 1.8364,
+      "step": 8190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.375,
+      "learning_rate": 1.882096082368644e-05,
+      "loss": 1.8843,
+      "step": 8191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.882067837175184e-05,
+      "loss": 1.7843,
+      "step": 8192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8820395888108987e-05,
+      "loss": 1.9256,
+      "step": 8193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8820113372758897e-05,
+      "loss": 1.7347,
+      "step": 8194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8819830825702582e-05,
+      "loss": 1.907,
+      "step": 8195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8819548246941068e-05,
+      "loss": 2.1342,
+      "step": 8196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8819265636475358e-05,
+      "loss": 1.9122,
+      "step": 8197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8818982994306476e-05,
+      "loss": 1.9302,
+      "step": 8198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8818700320435434e-05,
+      "loss": 2.1994,
+      "step": 8199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8818417614863252e-05,
+      "loss": 2.0014,
+      "step": 8200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8818134877590943e-05,
+      "loss": 1.969,
+      "step": 8201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8817852108619524e-05,
+      "loss": 1.8846,
+      "step": 8202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8817569307950015e-05,
+      "loss": 2.0264,
+      "step": 8203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8817286475583425e-05,
+      "loss": 2.0349,
+      "step": 8204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8817003611520777e-05,
+      "loss": 2.187,
+      "step": 8205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8816720715763088e-05,
+      "loss": 1.5651,
+      "step": 8206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.881643778831137e-05,
+      "loss": 1.4388,
+      "step": 8207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8816154829166646e-05,
+      "loss": 1.9427,
+      "step": 8208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8815871838329932e-05,
+      "loss": 1.5921,
+      "step": 8209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.881558881580224e-05,
+      "loss": 2.0212,
+      "step": 8210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8815305761584592e-05,
+      "loss": 2.8731,
+      "step": 8211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8815022675678005e-05,
+      "loss": 2.0813,
+      "step": 8212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8814739558083498e-05,
+      "loss": 1.9509,
+      "step": 8213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8814456408802085e-05,
+      "loss": 1.611,
+      "step": 8214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8814173227834783e-05,
+      "loss": 1.8891,
+      "step": 8215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8813890015182617e-05,
+      "loss": 2.2872,
+      "step": 8216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8813606770846598e-05,
+      "loss": 1.7392,
+      "step": 8217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.881332349482775e-05,
+      "loss": 1.9471,
+      "step": 8218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8813040187127085e-05,
+      "loss": 1.7131,
+      "step": 8219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8812756847745627e-05,
+      "loss": 1.832,
+      "step": 8220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8812473476684392e-05,
+      "loss": 2.211,
+      "step": 8221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8812190073944396e-05,
+      "loss": 2.0315,
+      "step": 8222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8811906639526665e-05,
+      "loss": 2.1814,
+      "step": 8223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.881162317343221e-05,
+      "loss": 1.8098,
+      "step": 8224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.881133967566206e-05,
+      "loss": 2.1009,
+      "step": 8225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.881105614621722e-05,
+      "loss": 2.0027,
+      "step": 8226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.375,
+      "learning_rate": 1.881077258509872e-05,
+      "loss": 2.1155,
+      "step": 8227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8810488992307578e-05,
+      "loss": 1.8141,
+      "step": 8228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.8810205367844808e-05,
+      "loss": 2.7853,
+      "step": 8229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8809921711711436e-05,
+      "loss": 1.8359,
+      "step": 8230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8809638023908475e-05,
+      "loss": 2.2718,
+      "step": 8231
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8809354304436954e-05,
+      "loss": 1.842,
+      "step": 8232
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8809070553297887e-05,
+      "loss": 1.904,
+      "step": 8233
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8808786770492292e-05,
+      "loss": 2.2048,
+      "step": 8234
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.880850295602119e-05,
+      "loss": 1.7406,
+      "step": 8235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8808219109885607e-05,
+      "loss": 1.8741,
+      "step": 8236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.880793523208656e-05,
+      "loss": 1.8588,
+      "step": 8237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8807651322625067e-05,
+      "loss": 1.8819,
+      "step": 8238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.880736738150215e-05,
+      "loss": 1.7239,
+      "step": 8239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.880708340871883e-05,
+      "loss": 1.9421,
+      "step": 8240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8806799404276132e-05,
+      "loss": 1.8385,
+      "step": 8241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.880651536817507e-05,
+      "loss": 1.8392,
+      "step": 8242
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8806231300416665e-05,
+      "loss": 2.0999,
+      "step": 8243
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8805947201001942e-05,
+      "loss": 1.9262,
+      "step": 8244
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8805663069931925e-05,
+      "loss": 1.6047,
+      "step": 8245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.880537890720763e-05,
+      "loss": 1.4389,
+      "step": 8246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8805094712830078e-05,
+      "loss": 1.9528,
+      "step": 8247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8804810486800295e-05,
+      "loss": 1.8229,
+      "step": 8248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8804526229119302e-05,
+      "loss": 2.1928,
+      "step": 8249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8804241939788118e-05,
+      "loss": 1.8065,
+      "step": 8250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8803957618807764e-05,
+      "loss": 2.1141,
+      "step": 8251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8803673266179267e-05,
+      "loss": 2.2061,
+      "step": 8252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8803388881903648e-05,
+      "loss": 1.7462,
+      "step": 8253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8803104465981925e-05,
+      "loss": 1.7754,
+      "step": 8254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8802820018415127e-05,
+      "loss": 1.8412,
+      "step": 8255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8802535539204268e-05,
+      "loss": 1.8702,
+      "step": 8256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.34375,
+      "learning_rate": 1.8802251028350376e-05,
+      "loss": 3.7829,
+      "step": 8257
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.880196648585448e-05,
+      "loss": 1.9278,
+      "step": 8258
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.880168191171759e-05,
+      "loss": 1.8059,
+      "step": 8259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8801397305940736e-05,
+      "loss": 1.6608,
+      "step": 8260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.880111266852494e-05,
+      "loss": 1.9211,
+      "step": 8261
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8800827999471226e-05,
+      "loss": 2.0761,
+      "step": 8262
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8800543298780615e-05,
+      "loss": 1.7794,
+      "step": 8263
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8800258566454133e-05,
+      "loss": 2.2404,
+      "step": 8264
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8799973802492803e-05,
+      "loss": 1.6287,
+      "step": 8265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.879968900689765e-05,
+      "loss": 1.9802,
+      "step": 8266
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.8799404179669694e-05,
+      "loss": 1.8323,
+      "step": 8267
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.879911932080996e-05,
+      "loss": 1.6134,
+      "step": 8268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8798834430319476e-05,
+      "loss": 1.9131,
+      "step": 8269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.879854950819926e-05,
+      "loss": 1.8862,
+      "step": 8270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8798264554450343e-05,
+      "loss": 1.8098,
+      "step": 8271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.8797979569073744e-05,
+      "loss": 2.0254,
+      "step": 8272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8797694552070486e-05,
+      "loss": 2.1569,
+      "step": 8273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.87974095034416e-05,
+      "loss": 1.7103,
+      "step": 8274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8797124423188105e-05,
+      "loss": 1.9989,
+      "step": 8275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.879683931131103e-05,
+      "loss": 2.1247,
+      "step": 8276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.87965541678114e-05,
+      "loss": 1.8222,
+      "step": 8277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8796268992690238e-05,
+      "loss": 1.6507,
+      "step": 8278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8795983785948568e-05,
+      "loss": 1.8581,
+      "step": 8279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.879569854758741e-05,
+      "loss": 1.6764,
+      "step": 8280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.8795413277607807e-05,
+      "loss": 1.6709,
+      "step": 8281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8795127976010765e-05,
+      "loss": 1.4643,
+      "step": 8282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8794842642797322e-05,
+      "loss": 1.8914,
+      "step": 8283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.87945572779685e-05,
+      "loss": 1.9867,
+      "step": 8284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.8794271881525323e-05,
+      "loss": 1.6655,
+      "step": 8285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8793986453468817e-05,
+      "loss": 1.9773,
+      "step": 8286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.879370099380001e-05,
+      "loss": 2.0553,
+      "step": 8287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.879341550251993e-05,
+      "loss": 2.2303,
+      "step": 8288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8793129979629596e-05,
+      "loss": 1.9474,
+      "step": 8289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8792844425130042e-05,
+      "loss": 1.8293,
+      "step": 8290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8792558839022294e-05,
+      "loss": 1.7545,
+      "step": 8291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8792273221307377e-05,
+      "loss": 1.8325,
+      "step": 8292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8791987571986313e-05,
+      "loss": 2.0341,
+      "step": 8293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8791701891060135e-05,
+      "loss": 2.0334,
+      "step": 8294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8791416178529868e-05,
+      "loss": 2.06,
+      "step": 8295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.8791130434396536e-05,
+      "loss": 1.8806,
+      "step": 8296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.879084465866117e-05,
+      "loss": 2.1279,
+      "step": 8297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.87905588513248e-05,
+      "loss": 1.9167,
+      "step": 8298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8790273012388448e-05,
+      "loss": 1.9941,
+      "step": 8299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8789987141853142e-05,
+      "loss": 2.3103,
+      "step": 8300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8789701239719913e-05,
+      "loss": 1.9991,
+      "step": 8301
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8789415305989786e-05,
+      "loss": 1.734,
+      "step": 8302
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.878912934066379e-05,
+      "loss": 1.8825,
+      "step": 8303
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.878884334374295e-05,
+      "loss": 1.8912,
+      "step": 8304
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.87885573152283e-05,
+      "loss": 2.2339,
+      "step": 8305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8788271255120864e-05,
+      "loss": 2.2609,
+      "step": 8306
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.878798516342167e-05,
+      "loss": 1.9938,
+      "step": 8307
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8787699040131747e-05,
+      "loss": 1.9957,
+      "step": 8308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8787412885252126e-05,
+      "loss": 1.6821,
+      "step": 8309
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8787126698783834e-05,
+      "loss": 1.7932,
+      "step": 8310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.87868404807279e-05,
+      "loss": 2.08,
+      "step": 8311
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.878655423108535e-05,
+      "loss": 1.7416,
+      "step": 8312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.878626794985722e-05,
+      "loss": 1.6173,
+      "step": 8313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.878598163704453e-05,
+      "loss": 1.9389,
+      "step": 8314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8785695292648316e-05,
+      "loss": 2.0003,
+      "step": 8315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8785408916669606e-05,
+      "loss": 1.5738,
+      "step": 8316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8785122509109425e-05,
+      "loss": 1.5236,
+      "step": 8317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.878483606996881e-05,
+      "loss": 2.1135,
+      "step": 8318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8784549599248783e-05,
+      "loss": 1.9765,
+      "step": 8319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8784263096950383e-05,
+      "loss": 2.2321,
+      "step": 8320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8783976563074628e-05,
+      "loss": 1.7703,
+      "step": 8321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8783689997622562e-05,
+      "loss": 1.8168,
+      "step": 8322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8783403400595205e-05,
+      "loss": 1.8932,
+      "step": 8323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8783116771993587e-05,
+      "loss": 1.9587,
+      "step": 8324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8782830111818745e-05,
+      "loss": 2.146,
+      "step": 8325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8782543420071702e-05,
+      "loss": 2.0641,
+      "step": 8326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8782256696753494e-05,
+      "loss": 1.8986,
+      "step": 8327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.878196994186515e-05,
+      "loss": 1.7316,
+      "step": 8328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8781683155407703e-05,
+      "loss": 1.9462,
+      "step": 8329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.878139633738218e-05,
+      "loss": 1.8363,
+      "step": 8330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8781109487789615e-05,
+      "loss": 1.9094,
+      "step": 8331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8780822606631036e-05,
+      "loss": 1.9161,
+      "step": 8332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.878053569390748e-05,
+      "loss": 1.6716,
+      "step": 8333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8780248749619972e-05,
+      "loss": 1.9066,
+      "step": 8334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8779961773769547e-05,
+      "loss": 1.8511,
+      "step": 8335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8779674766357235e-05,
+      "loss": 1.7629,
+      "step": 8336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.877938772738407e-05,
+      "loss": 2.2738,
+      "step": 8337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.877910065685108e-05,
+      "loss": 1.7237,
+      "step": 8338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.8778813554759302e-05,
+      "loss": 1.714,
+      "step": 8339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8778526421109763e-05,
+      "loss": 2.0911,
+      "step": 8340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.87782392559035e-05,
+      "loss": 1.8786,
+      "step": 8341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8777952059141542e-05,
+      "loss": 1.8777,
+      "step": 8342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.877766483082492e-05,
+      "loss": 2.0163,
+      "step": 8343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8777377570954672e-05,
+      "loss": 1.7285,
+      "step": 8344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8777090279531826e-05,
+      "loss": 1.8691,
+      "step": 8345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8776802956557416e-05,
+      "loss": 1.7141,
+      "step": 8346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8776515602032475e-05,
+      "loss": 1.8941,
+      "step": 8347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8776228215958038e-05,
+      "loss": 1.8925,
+      "step": 8348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8775940798335136e-05,
+      "loss": 1.6742,
+      "step": 8349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.87756533491648e-05,
+      "loss": 1.5771,
+      "step": 8350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8775365868448065e-05,
+      "loss": 1.8479,
+      "step": 8351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8775078356185966e-05,
+      "loss": 1.7656,
+      "step": 8352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8774790812379536e-05,
+      "loss": 1.7044,
+      "step": 8353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.877450323702981e-05,
+      "loss": 2.0243,
+      "step": 8354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8774215630137816e-05,
+      "loss": 1.9446,
+      "step": 8355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8773927991704596e-05,
+      "loss": 1.8238,
+      "step": 8356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8773640321731176e-05,
+      "loss": 1.8175,
+      "step": 8357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8773352620218596e-05,
+      "loss": 2.0534,
+      "step": 8358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.877306488716789e-05,
+      "loss": 1.8002,
+      "step": 8359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8772777122580085e-05,
+      "loss": 2.1172,
+      "step": 8360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8772489326456225e-05,
+      "loss": 1.8074,
+      "step": 8361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.877220149879734e-05,
+      "loss": 1.8209,
+      "step": 8362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8771913639604466e-05,
+      "loss": 2.04,
+      "step": 8363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8771625748878637e-05,
+      "loss": 2.4104,
+      "step": 8364
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.877133782662089e-05,
+      "loss": 2.0301,
+      "step": 8365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.877104987283225e-05,
+      "loss": 1.9349,
+      "step": 8366
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8770761887513766e-05,
+      "loss": 1.8264,
+      "step": 8367
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8770473870666466e-05,
+      "loss": 1.4972,
+      "step": 8368
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8770185822291387e-05,
+      "loss": 1.803,
+      "step": 8369
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.876989774238956e-05,
+      "loss": 1.5098,
+      "step": 8370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.876960963096203e-05,
+      "loss": 1.9894,
+      "step": 8371
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8769321488009827e-05,
+      "loss": 1.9592,
+      "step": 8372
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8769033313533985e-05,
+      "loss": 2.2847,
+      "step": 8373
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.876874510753554e-05,
+      "loss": 1.9147,
+      "step": 8374
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8768456870015538e-05,
+      "loss": 1.8508,
+      "step": 8375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8768168600975002e-05,
+      "loss": 1.8423,
+      "step": 8376
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8767880300414972e-05,
+      "loss": 2.1155,
+      "step": 8377
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.876759196833649e-05,
+      "loss": 1.788,
+      "step": 8378
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8767303604740586e-05,
+      "loss": 1.7067,
+      "step": 8379
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8767015209628297e-05,
+      "loss": 1.7934,
+      "step": 8380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.876672678300066e-05,
+      "loss": 1.9616,
+      "step": 8381
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.876643832485872e-05,
+      "loss": 2.0324,
+      "step": 8382
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8766149835203506e-05,
+      "loss": 1.7694,
+      "step": 8383
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8765861314036055e-05,
+      "loss": 1.9446,
+      "step": 8384
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8765572761357403e-05,
+      "loss": 1.9378,
+      "step": 8385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8765284177168594e-05,
+      "loss": 1.6333,
+      "step": 8386
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8764995561470657e-05,
+      "loss": 1.7351,
+      "step": 8387
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8764706914264636e-05,
+      "loss": 1.9224,
+      "step": 8388
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.876441823555157e-05,
+      "loss": 1.671,
+      "step": 8389
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.876412952533249e-05,
+      "loss": 1.9256,
+      "step": 8390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8763840783608436e-05,
+      "loss": 1.9471,
+      "step": 8391
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.876355201038045e-05,
+      "loss": 1.9133,
+      "step": 8392
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8763263205649563e-05,
+      "loss": 1.92,
+      "step": 8393
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8762974369416823e-05,
+      "loss": 1.4954,
+      "step": 8394
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8762685501683257e-05,
+      "loss": 1.838,
+      "step": 8395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.876239660244991e-05,
+      "loss": 1.8785,
+      "step": 8396
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8762107671717822e-05,
+      "loss": 1.6802,
+      "step": 8397
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8761818709488027e-05,
+      "loss": 1.6307,
+      "step": 8398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8761529715761566e-05,
+      "loss": 2.2351,
+      "step": 8399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.876124069053948e-05,
+      "loss": 2.1039,
+      "step": 8400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8760951633822803e-05,
+      "loss": 2.1309,
+      "step": 8401
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8760662545612577e-05,
+      "loss": 1.9232,
+      "step": 8402
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.876037342590984e-05,
+      "loss": 1.8961,
+      "step": 8403
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8760084274715634e-05,
+      "loss": 1.9502,
+      "step": 8404
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8759795092030996e-05,
+      "loss": 1.9378,
+      "step": 8405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8759505877856967e-05,
+      "loss": 1.8648,
+      "step": 8406
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8759216632194586e-05,
+      "loss": 1.9557,
+      "step": 8407
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8758927355044893e-05,
+      "loss": 1.8231,
+      "step": 8408
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8758638046408926e-05,
+      "loss": 1.5682,
+      "step": 8409
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8758348706287728e-05,
+      "loss": 1.8252,
+      "step": 8410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.875805933468234e-05,
+      "loss": 1.4774,
+      "step": 8411
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8757769931593795e-05,
+      "loss": 1.9533,
+      "step": 8412
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.875748049702314e-05,
+      "loss": 1.6283,
+      "step": 8413
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8757191030971413e-05,
+      "loss": 2.1148,
+      "step": 8414
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.875690153343966e-05,
+      "loss": 2.1268,
+      "step": 8415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.875661200442891e-05,
+      "loss": 2.2895,
+      "step": 8416
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8756322443940216e-05,
+      "loss": 2.0302,
+      "step": 8417
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.875603285197461e-05,
+      "loss": 1.9981,
+      "step": 8418
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.875574322853314e-05,
+      "loss": 1.9155,
+      "step": 8419
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8755453573616843e-05,
+      "loss": 2.2788,
+      "step": 8420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8755163887226756e-05,
+      "loss": 1.8403,
+      "step": 8421
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8754874169363934e-05,
+      "loss": 1.7219,
+      "step": 8422
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8754584420029403e-05,
+      "loss": 1.9146,
+      "step": 8423
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8754294639224216e-05,
+      "loss": 2.4056,
+      "step": 8424
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8754004826949404e-05,
+      "loss": 1.8607,
+      "step": 8425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.875371498320602e-05,
+      "loss": 1.9752,
+      "step": 8426
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.87534251079951e-05,
+      "loss": 1.9996,
+      "step": 8427
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8753135201317686e-05,
+      "loss": 2.0262,
+      "step": 8428
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.875284526317482e-05,
+      "loss": 2.0779,
+      "step": 8429
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8752555293567546e-05,
+      "loss": 2.0357,
+      "step": 8430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.875226529249691e-05,
+      "loss": 1.9755,
+      "step": 8431
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8751975259963947e-05,
+      "loss": 1.7478,
+      "step": 8432
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.87516851959697e-05,
+      "loss": 1.8904,
+      "step": 8433
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8751395100515216e-05,
+      "loss": 1.5196,
+      "step": 8434
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8751104973601538e-05,
+      "loss": 2.2464,
+      "step": 8435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8750814815229708e-05,
+      "loss": 2.0797,
+      "step": 8436
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8750524625400764e-05,
+      "loss": 1.8576,
+      "step": 8437
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8750234404115755e-05,
+      "loss": 1.6314,
+      "step": 8438
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8749944151375726e-05,
+      "loss": 2.0197,
+      "step": 8439
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8749653867181716e-05,
+      "loss": 2.3113,
+      "step": 8440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8749363551534766e-05,
+      "loss": 1.9817,
+      "step": 8441
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8749073204435924e-05,
+      "loss": 2.0367,
+      "step": 8442
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8748782825886234e-05,
+      "loss": 1.7536,
+      "step": 8443
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.874849241588674e-05,
+      "loss": 2.0238,
+      "step": 8444
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8748201974438484e-05,
+      "loss": 1.6139,
+      "step": 8445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.874791150154251e-05,
+      "loss": 1.9649,
+      "step": 8446
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8747620997199862e-05,
+      "loss": 2.0197,
+      "step": 8447
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8747330461411587e-05,
+      "loss": 2.091,
+      "step": 8448
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.874703989417873e-05,
+      "loss": 1.883,
+      "step": 8449
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.874674929550233e-05,
+      "loss": 1.9497,
+      "step": 8450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8746458665383434e-05,
+      "loss": 1.9596,
+      "step": 8451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.874616800382309e-05,
+      "loss": 1.8261,
+      "step": 8452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.874587731082234e-05,
+      "loss": 1.908,
+      "step": 8453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8745586586382228e-05,
+      "loss": 1.9125,
+      "step": 8454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8745295830503802e-05,
+      "loss": 1.7133,
+      "step": 8455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8745005043188104e-05,
+      "loss": 2.0101,
+      "step": 8456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8744714224436184e-05,
+      "loss": 1.8045,
+      "step": 8457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8744423374249083e-05,
+      "loss": 1.5361,
+      "step": 8458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8744132492627843e-05,
+      "loss": 1.6322,
+      "step": 8459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.8743841579573523e-05,
+      "loss": 2.087,
+      "step": 8460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8743550635087155e-05,
+      "loss": 2.1027,
+      "step": 8461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8743259659169793e-05,
+      "loss": 1.6418,
+      "step": 8462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8742968651822477e-05,
+      "loss": 1.8229,
+      "step": 8463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8742677613046258e-05,
+      "loss": 1.7603,
+      "step": 8464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8742386542842183e-05,
+      "loss": 1.7399,
+      "step": 8465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.874209544121129e-05,
+      "loss": 2.0713,
+      "step": 8466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.874180430815464e-05,
+      "loss": 1.9109,
+      "step": 8467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.8741513143673263e-05,
+      "loss": 1.9254,
+      "step": 8468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8741221947768218e-05,
+      "loss": 2.0656,
+      "step": 8469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8740930720440547e-05,
+      "loss": 1.9657,
+      "step": 8470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8740639461691295e-05,
+      "loss": 2.276,
+      "step": 8471
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.874034817152151e-05,
+      "loss": 2.0131,
+      "step": 8472
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8740056849932243e-05,
+      "loss": 2.0264,
+      "step": 8473
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.873976549692454e-05,
+      "loss": 1.8728,
+      "step": 8474
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8739474112499445e-05,
+      "loss": 2.1385,
+      "step": 8475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8739182696658008e-05,
+      "loss": 1.5636,
+      "step": 8476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8738891249401275e-05,
+      "loss": 1.826,
+      "step": 8477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8738599770730296e-05,
+      "loss": 1.7777,
+      "step": 8478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8738308260646117e-05,
+      "loss": 2.1056,
+      "step": 8479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8738016719149786e-05,
+      "loss": 2.1159,
+      "step": 8480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8737725146242353e-05,
+      "loss": 1.9736,
+      "step": 8481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.873743354192486e-05,
+      "loss": 1.7031,
+      "step": 8482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8737141906198364e-05,
+      "loss": 1.8205,
+      "step": 8483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.873685023906391e-05,
+      "loss": 1.9377,
+      "step": 8484
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.873655854052254e-05,
+      "loss": 1.8753,
+      "step": 8485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8736266810575313e-05,
+      "loss": 1.9629,
+      "step": 8486
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.873597504922327e-05,
+      "loss": 1.9905,
+      "step": 8487
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8735683256467466e-05,
+      "loss": 1.9777,
+      "step": 8488
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8735391432308946e-05,
+      "loss": 1.8004,
+      "step": 8489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8735099576748756e-05,
+      "loss": 1.8229,
+      "step": 8490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8734807689787953e-05,
+      "loss": 2.1753,
+      "step": 8491
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.873451577142758e-05,
+      "loss": 1.502,
+      "step": 8492
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.8734223821668685e-05,
+      "loss": 2.1262,
+      "step": 8493
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8733931840512327e-05,
+      "loss": 2.0271,
+      "step": 8494
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.873363982795955e-05,
+      "loss": 1.9435,
+      "step": 8495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.87333477840114e-05,
+      "loss": 1.8346,
+      "step": 8496
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8733055708668928e-05,
+      "loss": 2.118,
+      "step": 8497
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.873276360193319e-05,
+      "loss": 1.8467,
+      "step": 8498
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.873247146380523e-05,
+      "loss": 1.8314,
+      "step": 8499
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8732179294286103e-05,
+      "loss": 2.1974,
+      "step": 8500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8731887093376855e-05,
+      "loss": 1.8065,
+      "step": 8501
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8731594861078535e-05,
+      "loss": 1.9575,
+      "step": 8502
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.87313025973922e-05,
+      "loss": 1.8142,
+      "step": 8503
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8731010302318898e-05,
+      "loss": 1.9832,
+      "step": 8504
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8730717975859676e-05,
+      "loss": 2.2739,
+      "step": 8505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8730425618015585e-05,
+      "loss": 2.0567,
+      "step": 8506
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8730133228787684e-05,
+      "loss": 1.9085,
+      "step": 8507
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.872984080817702e-05,
+      "loss": 2.0734,
+      "step": 8508
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8729548356184634e-05,
+      "loss": 2.1282,
+      "step": 8509
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8729255872811593e-05,
+      "loss": 1.6841,
+      "step": 8510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.872896335805894e-05,
+      "loss": 1.8202,
+      "step": 8511
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8728670811927728e-05,
+      "loss": 1.9542,
+      "step": 8512
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8728378234419013e-05,
+      "loss": 2.0129,
+      "step": 8513
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8728085625533835e-05,
+      "loss": 1.7498,
+      "step": 8514
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8727792985273256e-05,
+      "loss": 2.1085,
+      "step": 8515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.872750031363833e-05,
+      "loss": 1.941,
+      "step": 8516
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.87272076106301e-05,
+      "loss": 1.8044,
+      "step": 8517
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.872691487624962e-05,
+      "loss": 2.07,
+      "step": 8518
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8726622110497946e-05,
+      "loss": 1.9092,
+      "step": 8519
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8726329313376132e-05,
+      "loss": 2.0209,
+      "step": 8520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.872603648488523e-05,
+      "loss": 1.6105,
+      "step": 8521
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8725743625026286e-05,
+      "loss": 1.7176,
+      "step": 8522
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.872545073380036e-05,
+      "loss": 1.799,
+      "step": 8523
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8725157811208503e-05,
+      "loss": 1.5718,
+      "step": 8524
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8724864857251762e-05,
+      "loss": 1.7522,
+      "step": 8525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8724571871931197e-05,
+      "loss": 1.9889,
+      "step": 8526
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8724278855247862e-05,
+      "loss": 1.7955,
+      "step": 8527
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8723985807202806e-05,
+      "loss": 1.9842,
+      "step": 8528
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8723692727797085e-05,
+      "loss": 1.993,
+      "step": 8529
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8723399617031754e-05,
+      "loss": 1.9157,
+      "step": 8530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8723106474907857e-05,
+      "loss": 1.9542,
+      "step": 8531
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.872281330142646e-05,
+      "loss": 2.0694,
+      "step": 8532
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8722520096588614e-05,
+      "loss": 1.9078,
+      "step": 8533
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8722226860395366e-05,
+      "loss": 2.0186,
+      "step": 8534
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.872193359284778e-05,
+      "loss": 1.8842,
+      "step": 8535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8721640293946904e-05,
+      "loss": 2.0654,
+      "step": 8536
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8721346963693795e-05,
+      "loss": 2.1176,
+      "step": 8537
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8721053602089502e-05,
+      "loss": 1.8963,
+      "step": 8538
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8720760209135087e-05,
+      "loss": 2.0626,
+      "step": 8539
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8720466784831603e-05,
+      "loss": 1.6856,
+      "step": 8540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8720173329180097e-05,
+      "loss": 1.9718,
+      "step": 8541
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8719879842181637e-05,
+      "loss": 1.6891,
+      "step": 8542
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8719586323837268e-05,
+      "loss": 2.1196,
+      "step": 8543
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.871929277414805e-05,
+      "loss": 2.1167,
+      "step": 8544
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8718999193115034e-05,
+      "loss": 1.8073,
+      "step": 8545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8718705580739278e-05,
+      "loss": 1.8765,
+      "step": 8546
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.871841193702184e-05,
+      "loss": 2.0637,
+      "step": 8547
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.871811826196377e-05,
+      "loss": 1.6857,
+      "step": 8548
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8717824555566126e-05,
+      "loss": 1.2937,
+      "step": 8549
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.871753081782997e-05,
+      "loss": 1.8722,
+      "step": 8550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8717237048756347e-05,
+      "loss": 1.8422,
+      "step": 8551
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.871694324834632e-05,
+      "loss": 1.6851,
+      "step": 8552
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8716649416600945e-05,
+      "loss": 1.8335,
+      "step": 8553
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8716355553521275e-05,
+      "loss": 1.8891,
+      "step": 8554
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8716061659108366e-05,
+      "loss": 1.9134,
+      "step": 8555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.871576773336328e-05,
+      "loss": 1.99,
+      "step": 8556
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.871547377628707e-05,
+      "loss": 2.219,
+      "step": 8557
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8715179787880794e-05,
+      "loss": 1.9576,
+      "step": 8558
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8714885768145504e-05,
+      "loss": 1.7415,
+      "step": 8559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.871459171708226e-05,
+      "loss": 1.7796,
+      "step": 8560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8714297634692124e-05,
+      "loss": 1.9637,
+      "step": 8561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8714003520976148e-05,
+      "loss": 1.8949,
+      "step": 8562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8713709375935392e-05,
+      "loss": 1.8938,
+      "step": 8563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8713415199570907e-05,
+      "loss": 1.852,
+      "step": 8564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8713120991883757e-05,
+      "loss": 2.2041,
+      "step": 8565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8712826752875e-05,
+      "loss": 1.7153,
+      "step": 8566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.871253248254569e-05,
+      "loss": 2.027,
+      "step": 8567
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8712238180896884e-05,
+      "loss": 1.7875,
+      "step": 8568
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.871194384792964e-05,
+      "loss": 1.692,
+      "step": 8569
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8711649483645025e-05,
+      "loss": 1.8208,
+      "step": 8570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8711355088044083e-05,
+      "loss": 2.19,
+      "step": 8571
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8711060661127886e-05,
+      "loss": 2.2808,
+      "step": 8572
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.871076620289748e-05,
+      "loss": 1.9285,
+      "step": 8573
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8710471713353934e-05,
+      "loss": 1.9736,
+      "step": 8574
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8710177192498303e-05,
+      "loss": 1.7769,
+      "step": 8575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8709882640331642e-05,
+      "loss": 1.6315,
+      "step": 8576
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8709588056855015e-05,
+      "loss": 1.8833,
+      "step": 8577
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8709293442069478e-05,
+      "loss": 1.7335,
+      "step": 8578
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.870899879597609e-05,
+      "loss": 2.025,
+      "step": 8579
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8708704118575912e-05,
+      "loss": 2.1787,
+      "step": 8580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.870840940987e-05,
+      "loss": 1.7415,
+      "step": 8581
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8708114669859418e-05,
+      "loss": 1.7196,
+      "step": 8582
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.870781989854522e-05,
+      "loss": 1.863,
+      "step": 8583
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8707525095928472e-05,
+      "loss": 2.2143,
+      "step": 8584
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.8707230262010233e-05,
+      "loss": 1.9236,
+      "step": 8585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8706935396791555e-05,
+      "loss": 1.7509,
+      "step": 8586
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8706640500273506e-05,
+      "loss": 1.9945,
+      "step": 8587
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.870634557245714e-05,
+      "loss": 1.8907,
+      "step": 8588
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8706050613343527e-05,
+      "loss": 1.8147,
+      "step": 8589
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8705755622933718e-05,
+      "loss": 1.7769,
+      "step": 8590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8705460601228773e-05,
+      "loss": 2.0874,
+      "step": 8591
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.870516554822976e-05,
+      "loss": 1.7443,
+      "step": 8592
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8704870463937735e-05,
+      "loss": 2.0674,
+      "step": 8593
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8704575348353757e-05,
+      "loss": 1.7579,
+      "step": 8594
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8704280201478894e-05,
+      "loss": 2.0639,
+      "step": 8595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8703985023314197e-05,
+      "loss": 1.9416,
+      "step": 8596
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8703689813860733e-05,
+      "loss": 1.9871,
+      "step": 8597
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8703394573119563e-05,
+      "loss": 1.9895,
+      "step": 8598
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.870309930109175e-05,
+      "loss": 2.0878,
+      "step": 8599
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8702803997778352e-05,
+      "loss": 1.9391,
+      "step": 8600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.8702508663180432e-05,
+      "loss": 2.1191,
+      "step": 8601
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.870221329729905e-05,
+      "loss": 1.8236,
+      "step": 8602
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.870191790013527e-05,
+      "loss": 2.1855,
+      "step": 8603
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8701622471690155e-05,
+      "loss": 2.0078,
+      "step": 8604
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8701327011964764e-05,
+      "loss": 1.7562,
+      "step": 8605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8701031520960158e-05,
+      "loss": 1.746,
+      "step": 8606
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8700735998677406e-05,
+      "loss": 1.9585,
+      "step": 8607
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8700440445117563e-05,
+      "loss": 2.0913,
+      "step": 8608
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8700144860281692e-05,
+      "loss": 2.0774,
+      "step": 8609
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.869984924417086e-05,
+      "loss": 1.7356,
+      "step": 8610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8699553596786127e-05,
+      "loss": 1.86,
+      "step": 8611
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8699257918128557e-05,
+      "loss": 1.9698,
+      "step": 8612
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8698962208199212e-05,
+      "loss": 2.0476,
+      "step": 8613
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8698666466999155e-05,
+      "loss": 1.7965,
+      "step": 8614
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8698370694529447e-05,
+      "loss": 1.8666,
+      "step": 8615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.869807489079116e-05,
+      "loss": 1.8932,
+      "step": 8616
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8697779055785344e-05,
+      "loss": 1.7484,
+      "step": 8617
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.869748318951307e-05,
+      "loss": 1.9743,
+      "step": 8618
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8697187291975403e-05,
+      "loss": 1.9168,
+      "step": 8619
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8696891363173405e-05,
+      "loss": 1.934,
+      "step": 8620
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8696595403108136e-05,
+      "loss": 1.9548,
+      "step": 8621
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8696299411780666e-05,
+      "loss": 1.9278,
+      "step": 8622
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8696003389192056e-05,
+      "loss": 1.8232,
+      "step": 8623
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.869570733534337e-05,
+      "loss": 1.8302,
+      "step": 8624
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.869541125023567e-05,
+      "loss": 1.8029,
+      "step": 8625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8695115133870024e-05,
+      "loss": 1.872,
+      "step": 8626
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.86948189862475e-05,
+      "loss": 1.9542,
+      "step": 8627
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8694522807369154e-05,
+      "loss": 2.3089,
+      "step": 8628
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.869422659723606e-05,
+      "loss": 2.0747,
+      "step": 8629
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.8693930355849267e-05,
+      "loss": 2.0681,
+      "step": 8630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.869363408320986e-05,
+      "loss": 1.8222,
+      "step": 8631
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.869333777931889e-05,
+      "loss": 1.853,
+      "step": 8632
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8693041444177428e-05,
+      "loss": 1.7664,
+      "step": 8633
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8692745077786537e-05,
+      "loss": 1.9248,
+      "step": 8634
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8692448680147282e-05,
+      "loss": 1.9807,
+      "step": 8635
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8692152251260732e-05,
+      "loss": 1.9542,
+      "step": 8636
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.869185579112795e-05,
+      "loss": 1.6109,
+      "step": 8637
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8691559299750005e-05,
+      "loss": 2.2479,
+      "step": 8638
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8691262777127953e-05,
+      "loss": 2.0055,
+      "step": 8639
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.869096622326287e-05,
+      "loss": 1.9029,
+      "step": 8640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8690669638155817e-05,
+      "loss": 1.663,
+      "step": 8641
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.869037302180787e-05,
+      "loss": 1.7848,
+      "step": 8642
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8690076374220078e-05,
+      "loss": 1.8945,
+      "step": 8643
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8689779695393522e-05,
+      "loss": 2.0452,
+      "step": 8644
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.868948298532926e-05,
+      "loss": 2.0648,
+      "step": 8645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.868918624402836e-05,
+      "loss": 1.6818,
+      "step": 8646
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8688889471491894e-05,
+      "loss": 1.908,
+      "step": 8647
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8688592667720925e-05,
+      "loss": 1.9225,
+      "step": 8648
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.868829583271652e-05,
+      "loss": 1.8406,
+      "step": 8649
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8687998966479744e-05,
+      "loss": 2.0194,
+      "step": 8650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8687702069011667e-05,
+      "loss": 1.8028,
+      "step": 8651
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8687405140313356e-05,
+      "loss": 1.8322,
+      "step": 8652
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.868710818038588e-05,
+      "loss": 1.5103,
+      "step": 8653
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8686811189230303e-05,
+      "loss": 1.8027,
+      "step": 8654
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8686514166847692e-05,
+      "loss": 2.5197,
+      "step": 8655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.868621711323912e-05,
+      "loss": 1.6879,
+      "step": 8656
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.868592002840565e-05,
+      "loss": 1.8455,
+      "step": 8657
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8685622912348352e-05,
+      "loss": 1.6795,
+      "step": 8658
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8685325765068297e-05,
+      "loss": 1.9683,
+      "step": 8659
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8685028586566546e-05,
+      "loss": 1.9263,
+      "step": 8660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.868473137684417e-05,
+      "loss": 1.8904,
+      "step": 8661
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8684434135902242e-05,
+      "loss": 1.885,
+      "step": 8662
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8684136863741825e-05,
+      "loss": 1.9739,
+      "step": 8663
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8683839560363993e-05,
+      "loss": 2.1243,
+      "step": 8664
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8683542225769805e-05,
+      "loss": 1.9283,
+      "step": 8665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8683244859960342e-05,
+      "loss": 1.5484,
+      "step": 8666
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8682947462936668e-05,
+      "loss": 1.8957,
+      "step": 8667
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8682650034699846e-05,
+      "loss": 1.5871,
+      "step": 8668
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8682352575250955e-05,
+      "loss": 1.8926,
+      "step": 8669
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8682055084591058e-05,
+      "loss": 1.6486,
+      "step": 8670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8681757562721226e-05,
+      "loss": 1.8326,
+      "step": 8671
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8681460009642533e-05,
+      "loss": 2.0574,
+      "step": 8672
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.868116242535604e-05,
+      "loss": 1.83,
+      "step": 8673
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8680864809862818e-05,
+      "loss": 1.7348,
+      "step": 8674
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8680567163163945e-05,
+      "loss": 1.7446,
+      "step": 8675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8680269485260486e-05,
+      "loss": 1.8985,
+      "step": 8676
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.867997177615351e-05,
+      "loss": 2.5801,
+      "step": 8677
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8679674035844085e-05,
+      "loss": 1.8128,
+      "step": 8678
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8679376264333288e-05,
+      "loss": 1.8297,
+      "step": 8679
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8679078461622184e-05,
+      "loss": 1.806,
+      "step": 8680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.867878062771185e-05,
+      "loss": 1.7502,
+      "step": 8681
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8678482762603347e-05,
+      "loss": 2.0498,
+      "step": 8682
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8678184866297753e-05,
+      "loss": 1.7705,
+      "step": 8683
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8677886938796137e-05,
+      "loss": 1.9314,
+      "step": 8684
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.867758898009957e-05,
+      "loss": 1.8083,
+      "step": 8685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8677290990209123e-05,
+      "loss": 2.0023,
+      "step": 8686
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.867699296912586e-05,
+      "loss": 1.9902,
+      "step": 8687
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.867669491685087e-05,
+      "loss": 1.9568,
+      "step": 8688
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.867639683338521e-05,
+      "loss": 2.0376,
+      "step": 8689
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.867609871872995e-05,
+      "loss": 1.7981,
+      "step": 8690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8675800572886175e-05,
+      "loss": 2.0505,
+      "step": 8691
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8675502395854942e-05,
+      "loss": 2.0286,
+      "step": 8692
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8675204187637335e-05,
+      "loss": 1.7282,
+      "step": 8693
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8674905948234418e-05,
+      "loss": 1.9572,
+      "step": 8694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8674607677647264e-05,
+      "loss": 2.5794,
+      "step": 8695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8674309375876952e-05,
+      "loss": 2.0286,
+      "step": 8696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8674011042924544e-05,
+      "loss": 1.8882,
+      "step": 8697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.867371267879112e-05,
+      "loss": 2.0192,
+      "step": 8698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8673414283477748e-05,
+      "loss": 1.9454,
+      "step": 8699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8673115856985507e-05,
+      "loss": 2.1668,
+      "step": 8700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8672817399315463e-05,
+      "loss": 2.0909,
+      "step": 8701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8672518910468692e-05,
+      "loss": 2.2708,
+      "step": 8702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8672220390446266e-05,
+      "loss": 1.7671,
+      "step": 8703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8671921839249257e-05,
+      "loss": 1.5255,
+      "step": 8704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8671623256878746e-05,
+      "loss": 1.7024,
+      "step": 8705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8671324643335795e-05,
+      "loss": 1.7327,
+      "step": 8706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.867102599862148e-05,
+      "loss": 1.8535,
+      "step": 8707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8670727322736884e-05,
+      "loss": 1.8042,
+      "step": 8708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.867042861568307e-05,
+      "loss": 1.7186,
+      "step": 8709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8670129877461117e-05,
+      "loss": 1.5443,
+      "step": 8710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8669831108072097e-05,
+      "loss": 2.0697,
+      "step": 8711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8669532307517082e-05,
+      "loss": 1.947,
+      "step": 8712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8669233475797154e-05,
+      "loss": 1.9238,
+      "step": 8713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.866893461291338e-05,
+      "loss": 2.0361,
+      "step": 8714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8668635718866832e-05,
+      "loss": 1.675,
+      "step": 8715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8668336793658593e-05,
+      "loss": 2.0018,
+      "step": 8716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.866803783728973e-05,
+      "loss": 2.0431,
+      "step": 8717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8667738849761323e-05,
+      "loss": 1.8111,
+      "step": 8718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.8667439831074447e-05,
+      "loss": 2.6651,
+      "step": 8719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8667140781230172e-05,
+      "loss": 2.0165,
+      "step": 8720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8666841700229575e-05,
+      "loss": 2.0655,
+      "step": 8721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.866654258807373e-05,
+      "loss": 2.0465,
+      "step": 8722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8666243444763718e-05,
+      "loss": 2.0072,
+      "step": 8723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8665944270300608e-05,
+      "loss": 1.7695,
+      "step": 8724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.866564506468548e-05,
+      "loss": 2.1311,
+      "step": 8725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8665345827919404e-05,
+      "loss": 1.9658,
+      "step": 8726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.866504656000346e-05,
+      "loss": 1.684,
+      "step": 8727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8664747260938724e-05,
+      "loss": 1.5652,
+      "step": 8728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8664447930726273e-05,
+      "loss": 1.979,
+      "step": 8729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8664148569367176e-05,
+      "loss": 1.5632,
+      "step": 8730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8663849176862515e-05,
+      "loss": 1.7436,
+      "step": 8731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.866354975321337e-05,
+      "loss": 1.9131,
+      "step": 8732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8663250298420808e-05,
+      "loss": 1.7319,
+      "step": 8733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8662950812485913e-05,
+      "loss": 2.0424,
+      "step": 8734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8662651295409753e-05,
+      "loss": 1.9353,
+      "step": 8735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8662351747193413e-05,
+      "loss": 1.7691,
+      "step": 8736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.866205216783797e-05,
+      "loss": 1.8884,
+      "step": 8737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8661752557344495e-05,
+      "loss": 1.82,
+      "step": 8738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8661452915714068e-05,
+      "loss": 2.2966,
+      "step": 8739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8661153242947767e-05,
+      "loss": 1.8442,
+      "step": 8740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.866085353904667e-05,
+      "loss": 1.7973,
+      "step": 8741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.866055380401185e-05,
+      "loss": 2.1422,
+      "step": 8742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.735,
+      "step": 8743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8659954240545358e-05,
+      "loss": 2.0255,
+      "step": 8744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8659654412115845e-05,
+      "loss": 1.693,
+      "step": 8745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8659354552556922e-05,
+      "loss": 1.8595,
+      "step": 8746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8659054661869668e-05,
+      "loss": 2.0892,
+      "step": 8747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.865875474005516e-05,
+      "loss": 2.0383,
+      "step": 8748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8658454787114472e-05,
+      "loss": 1.7822,
+      "step": 8749
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.865815480304869e-05,
+      "loss": 1.8633,
+      "step": 8750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8657854787858887e-05,
+      "loss": 1.8936,
+      "step": 8751
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8657554741546147e-05,
+      "loss": 2.0073,
+      "step": 8752
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8657254664111543e-05,
+      "loss": 1.7795,
+      "step": 8753
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8656954555556156e-05,
+      "loss": 2.0073,
+      "step": 8754
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8656654415881066e-05,
+      "loss": 1.8627,
+      "step": 8755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.865635424508735e-05,
+      "loss": 1.8473,
+      "step": 8756
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8656054043176085e-05,
+      "loss": 1.921,
+      "step": 8757
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8655753810148355e-05,
+      "loss": 2.0331,
+      "step": 8758
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8655453546005237e-05,
+      "loss": 1.7647,
+      "step": 8759
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8655153250747813e-05,
+      "loss": 2.1956,
+      "step": 8760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8654852924377157e-05,
+      "loss": 1.9237,
+      "step": 8761
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8654552566894352e-05,
+      "loss": 1.7028,
+      "step": 8762
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8654252178300477e-05,
+      "loss": 1.4638,
+      "step": 8763
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8653951758596613e-05,
+      "loss": 2.1946,
+      "step": 8764
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.865365130778384e-05,
+      "loss": 1.7559,
+      "step": 8765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8653350825863232e-05,
+      "loss": 1.6624,
+      "step": 8766
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.865305031283588e-05,
+      "loss": 1.9041,
+      "step": 8767
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8652749768702856e-05,
+      "loss": 2.159,
+      "step": 8768
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8652449193465244e-05,
+      "loss": 1.831,
+      "step": 8769
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8652148587124125e-05,
+      "loss": 1.937,
+      "step": 8770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8651847949680572e-05,
+      "loss": 1.7073,
+      "step": 8771
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8651547281135674e-05,
+      "loss": 1.6776,
+      "step": 8772
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.865124658149051e-05,
+      "loss": 1.8581,
+      "step": 8773
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8650945850746165e-05,
+      "loss": 1.6454,
+      "step": 8774
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.865064508890371e-05,
+      "loss": 1.976,
+      "step": 8775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8650344295964235e-05,
+      "loss": 1.6805,
+      "step": 8776
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8650043471928815e-05,
+      "loss": 1.9301,
+      "step": 8777
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.8649742616798537e-05,
+      "loss": 1.6868,
+      "step": 8778
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.864944173057448e-05,
+      "loss": 1.7636,
+      "step": 8779
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8649140813257725e-05,
+      "loss": 1.6246,
+      "step": 8780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8648839864849354e-05,
+      "loss": 1.8611,
+      "step": 8781
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8648538885350448e-05,
+      "loss": 1.7963,
+      "step": 8782
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8648237874762092e-05,
+      "loss": 2.3451,
+      "step": 8783
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8647936833085364e-05,
+      "loss": 2.0822,
+      "step": 8784
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.864763576032135e-05,
+      "loss": 1.6216,
+      "step": 8785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8647334656471127e-05,
+      "loss": 2.0225,
+      "step": 8786
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.8647033521535783e-05,
+      "loss": 2.1583,
+      "step": 8787
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.86467323555164e-05,
+      "loss": 1.8203,
+      "step": 8788
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.864643115841406e-05,
+      "loss": 1.7348,
+      "step": 8789
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.864612993022984e-05,
+      "loss": 1.9945,
+      "step": 8790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.864582867096483e-05,
+      "loss": 2.1953,
+      "step": 8791
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8645527380620113e-05,
+      "loss": 1.6201,
+      "step": 8792
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8645226059196768e-05,
+      "loss": 1.8231,
+      "step": 8793
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8644924706695882e-05,
+      "loss": 2.0112,
+      "step": 8794
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8644623323118533e-05,
+      "loss": 1.6802,
+      "step": 8795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.864432190846581e-05,
+      "loss": 1.7423,
+      "step": 8796
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8644020462738793e-05,
+      "loss": 2.1205,
+      "step": 8797
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8643718985938566e-05,
+      "loss": 1.9084,
+      "step": 8798
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8643417478066214e-05,
+      "loss": 1.9627,
+      "step": 8799
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.864311593912282e-05,
+      "loss": 1.8206,
+      "step": 8800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.864281436910947e-05,
+      "loss": 2.182,
+      "step": 8801
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8642512768027243e-05,
+      "loss": 2.0694,
+      "step": 8802
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8642211135877227e-05,
+      "loss": 1.8739,
+      "step": 8803
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8641909472660505e-05,
+      "loss": 1.6147,
+      "step": 8804
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8641607778378166e-05,
+      "loss": 1.9047,
+      "step": 8805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8641306053031286e-05,
+      "loss": 1.9468,
+      "step": 8806
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8641004296620958e-05,
+      "loss": 1.9944,
+      "step": 8807
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8640702509148264e-05,
+      "loss": 1.7825,
+      "step": 8808
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8640400690614285e-05,
+      "loss": 1.8679,
+      "step": 8809
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.864009884102011e-05,
+      "loss": 2.0885,
+      "step": 8810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8639796960366823e-05,
+      "loss": 1.5612,
+      "step": 8811
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.863949504865551e-05,
+      "loss": 2.1778,
+      "step": 8812
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8639193105887254e-05,
+      "loss": 1.8062,
+      "step": 8813
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.863889113206314e-05,
+      "loss": 2.0965,
+      "step": 8814
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.863858912718426e-05,
+      "loss": 1.6462,
+      "step": 8815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.863828709125169e-05,
+      "loss": 1.9357,
+      "step": 8816
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.863798502426653e-05,
+      "loss": 1.7467,
+      "step": 8817
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.863768292622985e-05,
+      "loss": 2.1972,
+      "step": 8818
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.863738079714274e-05,
+      "loss": 1.6857,
+      "step": 8819
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8637078637006294e-05,
+      "loss": 1.8061,
+      "step": 8820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.863677644582159e-05,
+      "loss": 2.0038,
+      "step": 8821
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8636474223589717e-05,
+      "loss": 1.8075,
+      "step": 8822
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8636171970311766e-05,
+      "loss": 1.8022,
+      "step": 8823
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8635869685988813e-05,
+      "loss": 1.6289,
+      "step": 8824
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8635567370621954e-05,
+      "loss": 1.7725,
+      "step": 8825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8635265024212273e-05,
+      "loss": 1.9703,
+      "step": 8826
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8634962646760856e-05,
+      "loss": 1.7825,
+      "step": 8827
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.863466023826879e-05,
+      "loss": 2.1706,
+      "step": 8828
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8634357798737163e-05,
+      "loss": 1.9147,
+      "step": 8829
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.863405532816706e-05,
+      "loss": 1.7935,
+      "step": 8830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.863375282655957e-05,
+      "loss": 1.6966,
+      "step": 8831
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.863345029391578e-05,
+      "loss": 1.7338,
+      "step": 8832
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8633147730236782e-05,
+      "loss": 2.0804,
+      "step": 8833
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8632845135523657e-05,
+      "loss": 2.9934,
+      "step": 8834
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8632542509777492e-05,
+      "loss": 1.6879,
+      "step": 8835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8632239852999384e-05,
+      "loss": 1.9416,
+      "step": 8836
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8631937165190413e-05,
+      "loss": 2.034,
+      "step": 8837
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.863163444635167e-05,
+      "loss": 2.1324,
+      "step": 8838
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8631331696484242e-05,
+      "loss": 1.531,
+      "step": 8839
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8631028915589217e-05,
+      "loss": 1.8052,
+      "step": 8840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8630726103667684e-05,
+      "loss": 1.9742,
+      "step": 8841
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8630423260720732e-05,
+      "loss": 1.7872,
+      "step": 8842
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.863012038674945e-05,
+      "loss": 1.8076,
+      "step": 8843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.8629817481754928e-05,
+      "loss": 1.7914,
+      "step": 8844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.862951454573825e-05,
+      "loss": 1.6708,
+      "step": 8845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.862921157870051e-05,
+      "loss": 2.1026,
+      "step": 8846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8628908580642795e-05,
+      "loss": 1.8825,
+      "step": 8847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8628605551566192e-05,
+      "loss": 1.908,
+      "step": 8848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.8628302491471794e-05,
+      "loss": 1.65,
+      "step": 8849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.862799940036069e-05,
+      "loss": 1.9233,
+      "step": 8850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8627696278233966e-05,
+      "loss": 1.5385,
+      "step": 8851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8627393125092717e-05,
+      "loss": 2.1412,
+      "step": 8852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.862708994093803e-05,
+      "loss": 1.9785,
+      "step": 8853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8626786725770993e-05,
+      "loss": 1.8535,
+      "step": 8854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.86264834795927e-05,
+      "loss": 1.6348,
+      "step": 8855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.862618020240424e-05,
+      "loss": 2.098,
+      "step": 8856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8625876894206697e-05,
+      "loss": 2.068,
+      "step": 8857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.862557355500117e-05,
+      "loss": 1.7667,
+      "step": 8858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8625270184788746e-05,
+      "loss": 2.0834,
+      "step": 8859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8624966783570517e-05,
+      "loss": 1.9413,
+      "step": 8860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.862466335134757e-05,
+      "loss": 1.847,
+      "step": 8861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8624359888120996e-05,
+      "loss": 2.079,
+      "step": 8862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8624056393891892e-05,
+      "loss": 2.125,
+      "step": 8863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8623752868661342e-05,
+      "loss": 1.8991,
+      "step": 8864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8623449312430443e-05,
+      "loss": 1.7238,
+      "step": 8865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.862314572520028e-05,
+      "loss": 1.8237,
+      "step": 8866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.862284210697195e-05,
+      "loss": 1.9398,
+      "step": 8867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.862253845774654e-05,
+      "loss": 2.2838,
+      "step": 8868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8622234777525145e-05,
+      "loss": 1.7477,
+      "step": 8869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8621931066308853e-05,
+      "loss": 1.8486,
+      "step": 8870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8621627324098758e-05,
+      "loss": 1.9714,
+      "step": 8871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.862132355089595e-05,
+      "loss": 1.7693,
+      "step": 8872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8621019746701524e-05,
+      "loss": 1.6738,
+      "step": 8873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8620715911516573e-05,
+      "loss": 1.8649,
+      "step": 8874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8620412045342187e-05,
+      "loss": 1.9324,
+      "step": 8875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8620108148179456e-05,
+      "loss": 1.9094,
+      "step": 8876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8619804220029475e-05,
+      "loss": 1.7752,
+      "step": 8877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8619500260893338e-05,
+      "loss": 2.0101,
+      "step": 8878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.375,
+      "learning_rate": 1.861919627077214e-05,
+      "loss": 1.4652,
+      "step": 8879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8618892249666963e-05,
+      "loss": 1.6146,
+      "step": 8880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8618588197578907e-05,
+      "loss": 2.0458,
+      "step": 8881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8618284114509067e-05,
+      "loss": 1.6469,
+      "step": 8882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8617980000458532e-05,
+      "loss": 1.6643,
+      "step": 8883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.86176758554284e-05,
+      "loss": 2.5768,
+      "step": 8884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8617371679419757e-05,
+      "loss": 1.9897,
+      "step": 8885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8617067472433705e-05,
+      "loss": 2.0686,
+      "step": 8886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8616763234471334e-05,
+      "loss": 1.7827,
+      "step": 8887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8616458965533733e-05,
+      "loss": 1.83,
+      "step": 8888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8616154665622e-05,
+      "loss": 1.6793,
+      "step": 8889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.861585033473723e-05,
+      "loss": 1.5754,
+      "step": 8890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8615545972880516e-05,
+      "loss": 1.9482,
+      "step": 8891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8615241580052953e-05,
+      "loss": 1.9987,
+      "step": 8892
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8614937156255632e-05,
+      "loss": 2.0954,
+      "step": 8893
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.861463270148965e-05,
+      "loss": 1.5374,
+      "step": 8894
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8614328215756102e-05,
+      "loss": 2.6428,
+      "step": 8895
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8614023699056076e-05,
+      "loss": 1.7259,
+      "step": 8896
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8613719151390676e-05,
+      "loss": 2.0269,
+      "step": 8897
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8613414572760994e-05,
+      "loss": 1.6763,
+      "step": 8898
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8613109963168125e-05,
+      "loss": 2.1226,
+      "step": 8899
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8612805322613158e-05,
+      "loss": 1.8885,
+      "step": 8900
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.8612500651097193e-05,
+      "loss": 2.3984,
+      "step": 8901
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8612195948621326e-05,
+      "loss": 1.8184,
+      "step": 8902
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8611891215186653e-05,
+      "loss": 1.7519,
+      "step": 8903
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8611586450794267e-05,
+      "loss": 1.8615,
+      "step": 8904
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8611281655445262e-05,
+      "loss": 1.8946,
+      "step": 8905
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8610976829140738e-05,
+      "loss": 1.878,
+      "step": 8906
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8610671971881787e-05,
+      "loss": 2.0142,
+      "step": 8907
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8610367083669508e-05,
+      "loss": 1.7811,
+      "step": 8908
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8610062164504995e-05,
+      "loss": 2.1217,
+      "step": 8909
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8609757214389346e-05,
+      "loss": 1.7529,
+      "step": 8910
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8609452233323656e-05,
+      "loss": 1.7374,
+      "step": 8911
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.860914722130902e-05,
+      "loss": 1.9474,
+      "step": 8912
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8608842178346536e-05,
+      "loss": 2.1887,
+      "step": 8913
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8608537104437303e-05,
+      "loss": 1.9671,
+      "step": 8914
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8608231999582413e-05,
+      "loss": 1.6243,
+      "step": 8915
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8607926863782966e-05,
+      "loss": 1.8961,
+      "step": 8916
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8607621697040053e-05,
+      "loss": 1.9022,
+      "step": 8917
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.860731649935478e-05,
+      "loss": 1.6988,
+      "step": 8918
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8607011270728237e-05,
+      "loss": 1.9494,
+      "step": 8919
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8606706011161525e-05,
+      "loss": 2.184,
+      "step": 8920
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8606400720655742e-05,
+      "loss": 2.0736,
+      "step": 8921
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8606095399211982e-05,
+      "loss": 1.8001,
+      "step": 8922
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8605790046831344e-05,
+      "loss": 1.6123,
+      "step": 8923
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.8605484663514927e-05,
+      "loss": 2.341,
+      "step": 8924
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.860517924926383e-05,
+      "loss": 1.9053,
+      "step": 8925
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8604873804079146e-05,
+      "loss": 1.5912,
+      "step": 8926
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8604568327961976e-05,
+      "loss": 1.6908,
+      "step": 8927
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8604262820913416e-05,
+      "loss": 1.8144,
+      "step": 8928
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8603957282934567e-05,
+      "loss": 1.9218,
+      "step": 8929
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.860365171402653e-05,
+      "loss": 2.0023,
+      "step": 8930
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8603346114190393e-05,
+      "loss": 2.0115,
+      "step": 8931
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8603040483427267e-05,
+      "loss": 1.7607,
+      "step": 8932
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8602734821738243e-05,
+      "loss": 2.008,
+      "step": 8933
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8602429129124422e-05,
+      "loss": 1.93,
+      "step": 8934
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8602123405586906e-05,
+      "loss": 2.0849,
+      "step": 8935
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8601817651126785e-05,
+      "loss": 1.8834,
+      "step": 8936
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8601511865745167e-05,
+      "loss": 1.7304,
+      "step": 8937
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8601206049443147e-05,
+      "loss": 1.7531,
+      "step": 8938
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.860090020222183e-05,
+      "loss": 2.0373,
+      "step": 8939
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8600594324082303e-05,
+      "loss": 1.889,
+      "step": 8940
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8600288415025676e-05,
+      "loss": 1.678,
+      "step": 8941
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.859998247505305e-05,
+      "loss": 2.0212,
+      "step": 8942
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.8599676504165515e-05,
+      "loss": 1.9921,
+      "step": 8943
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.859937050236418e-05,
+      "loss": 1.9994,
+      "step": 8944
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.859906446965014e-05,
+      "loss": 1.8853,
+      "step": 8945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.85987584060245e-05,
+      "loss": 2.0453,
+      "step": 8946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8598452311488354e-05,
+      "loss": 2.0314,
+      "step": 8947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8598146186042808e-05,
+      "loss": 1.8709,
+      "step": 8948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8597840029688958e-05,
+      "loss": 1.8891,
+      "step": 8949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8597533842427907e-05,
+      "loss": 1.7335,
+      "step": 8950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8597227624260754e-05,
+      "loss": 2.6228,
+      "step": 8951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8596921375188605e-05,
+      "loss": 1.8814,
+      "step": 8952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8596615095212553e-05,
+      "loss": 1.8966,
+      "step": 8953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8596308784333704e-05,
+      "loss": 1.6763,
+      "step": 8954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8596002442553155e-05,
+      "loss": 1.9735,
+      "step": 8955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8595696069872013e-05,
+      "loss": 1.9518,
+      "step": 8956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8595389666291378e-05,
+      "loss": 2.384,
+      "step": 8957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8595083231812346e-05,
+      "loss": 1.9711,
+      "step": 8958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8594776766436025e-05,
+      "loss": 1.9076,
+      "step": 8959
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8594470270163514e-05,
+      "loss": 2.2117,
+      "step": 8960
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8594163742995913e-05,
+      "loss": 1.8278,
+      "step": 8961
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8593857184934327e-05,
+      "loss": 1.3614,
+      "step": 8962
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.8593550595979857e-05,
+      "loss": 1.6423,
+      "step": 8963
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8593243976133607e-05,
+      "loss": 2.1239,
+      "step": 8964
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8592937325396672e-05,
+      "loss": 1.8162,
+      "step": 8965
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.859263064377016e-05,
+      "loss": 1.9474,
+      "step": 8966
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.859232393125518e-05,
+      "loss": 1.7604,
+      "step": 8967
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8592017187852823e-05,
+      "loss": 1.8233,
+      "step": 8968
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8591710413564193e-05,
+      "loss": 1.8464,
+      "step": 8969
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.85914036083904e-05,
+      "loss": 1.7392,
+      "step": 8970
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.859109677233254e-05,
+      "loss": 2.1622,
+      "step": 8971
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8590789905391723e-05,
+      "loss": 1.8872,
+      "step": 8972
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8590483007569047e-05,
+      "loss": 1.8633,
+      "step": 8973
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8590176078865613e-05,
+      "loss": 1.9284,
+      "step": 8974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8589869119282532e-05,
+      "loss": 2.0063,
+      "step": 8975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.85895621288209e-05,
+      "loss": 1.9955,
+      "step": 8976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8589255107481822e-05,
+      "loss": 2.1572,
+      "step": 8977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8588948055266405e-05,
+      "loss": 1.9011,
+      "step": 8978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.858864097217575e-05,
+      "loss": 1.6953,
+      "step": 8979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8588333858210966e-05,
+      "loss": 1.9109,
+      "step": 8980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8588026713373147e-05,
+      "loss": 1.7993,
+      "step": 8981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8587719537663406e-05,
+      "loss": 1.9669,
+      "step": 8982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8587412331082845e-05,
+      "loss": 2.2171,
+      "step": 8983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8587105093632564e-05,
+      "loss": 2.2439,
+      "step": 8984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8586797825313675e-05,
+      "loss": 1.6976,
+      "step": 8985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.858649052612728e-05,
+      "loss": 2.0253,
+      "step": 8986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8586183196074475e-05,
+      "loss": 1.7177,
+      "step": 8987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8585875835156376e-05,
+      "loss": 1.9335,
+      "step": 8988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8585568443374087e-05,
+      "loss": 1.9511,
+      "step": 8989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8585261020728705e-05,
+      "loss": 1.7239,
+      "step": 8990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8584953567221343e-05,
+      "loss": 1.6541,
+      "step": 8991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8584646082853104e-05,
+      "loss": 1.801,
+      "step": 8992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8584338567625087e-05,
+      "loss": 2.0008,
+      "step": 8993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.858403102153841e-05,
+      "loss": 2.035,
+      "step": 8994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.858372344459417e-05,
+      "loss": 2.0037,
+      "step": 8995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.858341583679347e-05,
+      "loss": 1.9626,
+      "step": 8996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8583108198137422e-05,
+      "loss": 1.6932,
+      "step": 8997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.858280052862713e-05,
+      "loss": 1.8835,
+      "step": 8998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.8582492828263702e-05,
+      "loss": 1.9871,
+      "step": 8999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8582185097048236e-05,
+      "loss": 1.7543,
+      "step": 9000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8581877334981852e-05,
+      "loss": 2.0432,
+      "step": 9001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8581569542065643e-05,
+      "loss": 1.7784,
+      "step": 9002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8581261718300722e-05,
+      "loss": 1.9609,
+      "step": 9003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.85809538636882e-05,
+      "loss": 1.628,
+      "step": 9004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8580645978229172e-05,
+      "loss": 1.8864,
+      "step": 9005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.858033806192475e-05,
+      "loss": 1.676,
+      "step": 9006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8580030114776045e-05,
+      "loss": 2.2698,
+      "step": 9007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.857972213678416e-05,
+      "loss": 1.7111,
+      "step": 9008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8579414127950204e-05,
+      "loss": 1.6474,
+      "step": 9009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.8579106088275283e-05,
+      "loss": 2.3798,
+      "step": 9010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8578798017760502e-05,
+      "loss": 1.8796,
+      "step": 9011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8578489916406973e-05,
+      "loss": 1.9069,
+      "step": 9012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8578181784215804e-05,
+      "loss": 1.6804,
+      "step": 9013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8577873621188095e-05,
+      "loss": 1.8095,
+      "step": 9014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.857756542732496e-05,
+      "loss": 2.4387,
+      "step": 9015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.857725720262751e-05,
+      "loss": 1.8183,
+      "step": 9016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8576948947096845e-05,
+      "loss": 2.0391,
+      "step": 9017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.857664066073408e-05,
+      "loss": 1.8042,
+      "step": 9018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.857633234354032e-05,
+      "loss": 2.0207,
+      "step": 9019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.857602399551667e-05,
+      "loss": 2.1741,
+      "step": 9020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.857571561666424e-05,
+      "loss": 2.277,
+      "step": 9021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8575407206984147e-05,
+      "loss": 1.6956,
+      "step": 9022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.857509876647749e-05,
+      "loss": 2.0421,
+      "step": 9023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.857479029514538e-05,
+      "loss": 1.8496,
+      "step": 9024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.857448179298893e-05,
+      "loss": 2.0293,
+      "step": 9025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8574173260009244e-05,
+      "loss": 2.0565,
+      "step": 9026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.857386469620743e-05,
+      "loss": 2.5264,
+      "step": 9027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8573556101584605e-05,
+      "loss": 1.869,
+      "step": 9028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.857324747614187e-05,
+      "loss": 1.694,
+      "step": 9029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8572938819880337e-05,
+      "loss": 1.6542,
+      "step": 9030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.857263013280112e-05,
+      "loss": 2.0435,
+      "step": 9031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8572321414905322e-05,
+      "loss": 2.0135,
+      "step": 9032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8572012666194056e-05,
+      "loss": 2.1155,
+      "step": 9033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8571703886668433e-05,
+      "loss": 1.9189,
+      "step": 9034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8571395076329564e-05,
+      "loss": 1.752,
+      "step": 9035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8571086235178553e-05,
+      "loss": 1.9972,
+      "step": 9036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8570777363216515e-05,
+      "loss": 1.6874,
+      "step": 9037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.857046846044456e-05,
+      "loss": 1.7784,
+      "step": 9038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8570159526863795e-05,
+      "loss": 1.8006,
+      "step": 9039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.856985056247534e-05,
+      "loss": 1.9167,
+      "step": 9040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.856954156728029e-05,
+      "loss": 1.7846,
+      "step": 9041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8569232541279772e-05,
+      "loss": 1.778,
+      "step": 9042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8568923484474884e-05,
+      "loss": 1.9616,
+      "step": 9043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.8568614396866746e-05,
+      "loss": 2.1906,
+      "step": 9044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8568305278456464e-05,
+      "loss": 2.3188,
+      "step": 9045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8567996129245154e-05,
+      "loss": 1.8817,
+      "step": 9046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.856768694923392e-05,
+      "loss": 1.8361,
+      "step": 9047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.856737773842388e-05,
+      "loss": 2.2172,
+      "step": 9048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8567068496816143e-05,
+      "loss": 2.2232,
+      "step": 9049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8566759224411822e-05,
+      "loss": 1.9225,
+      "step": 9050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8566449921212023e-05,
+      "loss": 2.029,
+      "step": 9051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8566140587217867e-05,
+      "loss": 1.8158,
+      "step": 9052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.856583122243046e-05,
+      "loss": 2.0949,
+      "step": 9053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.8565521826850915e-05,
+      "loss": 2.4054,
+      "step": 9054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8565212400480345e-05,
+      "loss": 1.8431,
+      "step": 9055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.8564902943319866e-05,
+      "loss": 2.147,
+      "step": 9056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.856459345537058e-05,
+      "loss": 2.184,
+      "step": 9057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8564283936633608e-05,
+      "loss": 1.821,
+      "step": 9058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8563974387110062e-05,
+      "loss": 2.1864,
+      "step": 9059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8563664806801052e-05,
+      "loss": 2.0015,
+      "step": 9060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8563355195707694e-05,
+      "loss": 1.736,
+      "step": 9061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.85630455538311e-05,
+      "loss": 2.0587,
+      "step": 9062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.856273588117238e-05,
+      "loss": 1.7409,
+      "step": 9063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.856242617773265e-05,
+      "loss": 1.7286,
+      "step": 9064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8562116443513024e-05,
+      "loss": 2.4243,
+      "step": 9065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.856180667851461e-05,
+      "loss": 1.9823,
+      "step": 9066
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.856149688273853e-05,
+      "loss": 1.8937,
+      "step": 9067
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8561187056185893e-05,
+      "loss": 1.9827,
+      "step": 9068
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8560877198857814e-05,
+      "loss": 1.9368,
+      "step": 9069
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8560567310755405e-05,
+      "loss": 1.757,
+      "step": 9070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8560257391879778e-05,
+      "loss": 1.9681,
+      "step": 9071
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8559947442232053e-05,
+      "loss": 1.4528,
+      "step": 9072
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8559637461813343e-05,
+      "loss": 1.8857,
+      "step": 9073
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.855932745062476e-05,
+      "loss": 1.7036,
+      "step": 9074
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8559017408667416e-05,
+      "loss": 1.8698,
+      "step": 9075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8558707335942433e-05,
+      "loss": 1.965,
+      "step": 9076
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.8558397232450916e-05,
+      "loss": 2.7565,
+      "step": 9077
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8558087098193987e-05,
+      "loss": 1.5975,
+      "step": 9078
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.855777693317276e-05,
+      "loss": 1.8553,
+      "step": 9079
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.855746673738835e-05,
+      "loss": 1.9899,
+      "step": 9080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8557156510841866e-05,
+      "loss": 1.773,
+      "step": 9081
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.8556846253534434e-05,
+      "loss": 2.0759,
+      "step": 9082
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.855653596546716e-05,
+      "loss": 1.8922,
+      "step": 9083
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8556225646641163e-05,
+      "loss": 2.0114,
+      "step": 9084
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.855591529705756e-05,
+      "loss": 2.0396,
+      "step": 9085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8555604916717463e-05,
+      "loss": 1.6423,
+      "step": 9086
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.855529450562199e-05,
+      "loss": 1.6545,
+      "step": 9087
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.8554984063772257e-05,
+      "loss": 1.7579,
+      "step": 9088
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.875,
+      "learning_rate": 1.855467359116938e-05,
+      "loss": 2.0042,
+      "step": 9089
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8554363087814472e-05,
+      "loss": 1.9089,
+      "step": 9090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.8554052553708654e-05,
+      "loss": 1.7291,
+      "step": 9091
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.855374198885304e-05,
+      "loss": 1.63,
+      "step": 9092
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8553431393248743e-05,
+      "loss": 1.9063,
+      "step": 9093
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.855312076689689e-05,
+      "loss": 2.1543,
+      "step": 9094
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8552810109798584e-05,
+      "loss": 1.9585,
+      "step": 9095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.855249942195495e-05,
+      "loss": 1.7871,
+      "step": 9096
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8552188703367105e-05,
+      "loss": 2.1357,
+      "step": 9097
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8551877954036165e-05,
+      "loss": 1.9523,
+      "step": 9098
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8551567173963244e-05,
+      "loss": 1.6583,
+      "step": 9099
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.855125636314946e-05,
+      "loss": 1.7688,
+      "step": 9100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8550945521595935e-05,
+      "loss": 1.6182,
+      "step": 9101
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.855063464930378e-05,
+      "loss": 1.6944,
+      "step": 9102
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.855032374627412e-05,
+      "loss": 1.7789,
+      "step": 9103
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8550012812508064e-05,
+      "loss": 2.0551,
+      "step": 9104
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8549701848006733e-05,
+      "loss": 1.9829,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8549390852771248e-05,
+      "loss": 1.7957,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.854907982680273e-05,
+      "loss": 1.7979,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.8548768770102284e-05,
+      "loss": 2.7843,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8548457682671038e-05,
+      "loss": 1.8878,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.854814656451011e-05,
+      "loss": 2.0048,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8547835415620617e-05,
+      "loss": 1.8048,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8547524236003675e-05,
+      "loss": 1.9008,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8547213025660406e-05,
+      "loss": 2.0509,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8546901784591928e-05,
+      "loss": 2.0008,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.854659051279936e-05,
+      "loss": 1.7584,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8546279210283815e-05,
+      "loss": 1.9905,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8545967877046423e-05,
+      "loss": 1.9394,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8545656513088298e-05,
+      "loss": 1.8449,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8545345118410554e-05,
+      "loss": 2.0555,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8545033693014317e-05,
+      "loss": 1.9267,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8544722236900708e-05,
+      "loss": 1.9325,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.854441075007084e-05,
+      "loss": 1.8383,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8544099232525836e-05,
+      "loss": 1.6648,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8543787684266816e-05,
+      "loss": 2.0448,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8543476105294903e-05,
+      "loss": 1.8032,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8543164495611212e-05,
+      "loss": 2.1057,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8542852855216865e-05,
+      "loss": 1.6517,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8542541184112982e-05,
+      "loss": 1.8678,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.854222948230068e-05,
+      "loss": 2.0441,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.854191774978109e-05,
+      "loss": 1.5998,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8541605986555322e-05,
+      "loss": 2.1105,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8541294192624503e-05,
+      "loss": 2.1133,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.854098236798975e-05,
+      "loss": 2.0955,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.854067051265218e-05,
+      "loss": 1.8448,
+      "step": 9134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8540358626612926e-05,
+      "loss": 1.9797,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.85400467098731e-05,
+      "loss": 1.7451,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8539734762433825e-05,
+      "loss": 1.5017,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8539422784296226e-05,
+      "loss": 1.6421,
+      "step": 9138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8539110775461422e-05,
+      "loss": 2.008,
+      "step": 9139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8538798735930533e-05,
+      "loss": 1.6249,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.853848666570468e-05,
+      "loss": 1.5893,
+      "step": 9141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8538174564784985e-05,
+      "loss": 1.5647,
+      "step": 9142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8537862433172575e-05,
+      "loss": 2.0018,
+      "step": 9143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.853755027086857e-05,
+      "loss": 2.137,
+      "step": 9144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8537238077874088e-05,
+      "loss": 2.0983,
+      "step": 9145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8536925854190254e-05,
+      "loss": 1.5145,
+      "step": 9146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8536613599818193e-05,
+      "loss": 1.7023,
+      "step": 9147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8536301314759026e-05,
+      "loss": 1.9176,
+      "step": 9148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.853598899901387e-05,
+      "loss": 1.904,
+      "step": 9149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8535676652583853e-05,
+      "loss": 1.6093,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.85353642754701e-05,
+      "loss": 1.9313,
+      "step": 9151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.853505186767373e-05,
+      "loss": 1.5932,
+      "step": 9152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8534739429195867e-05,
+      "loss": 1.9882,
+      "step": 9153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8534426960037633e-05,
+      "loss": 2.0378,
+      "step": 9154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8534114460200154e-05,
+      "loss": 1.6026,
+      "step": 9155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.853380192968455e-05,
+      "loss": 1.8235,
+      "step": 9156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8533489368491947e-05,
+      "loss": 1.9596,
+      "step": 9157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8533176776623467e-05,
+      "loss": 1.891,
+      "step": 9158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8532864154080235e-05,
+      "loss": 1.8602,
+      "step": 9159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8532551500863377e-05,
+      "loss": 1.9845,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.853223881697401e-05,
+      "loss": 2.0095,
+      "step": 9161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8531926102413265e-05,
+      "loss": 2.0741,
+      "step": 9162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8531613357182264e-05,
+      "loss": 1.8292,
+      "step": 9163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8531300581282127e-05,
+      "loss": 2.0431,
+      "step": 9164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8530987774713986e-05,
+      "loss": 1.9524,
+      "step": 9165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.853067493747896e-05,
+      "loss": 1.7486,
+      "step": 9166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8530362069578174e-05,
+      "loss": 1.9256,
+      "step": 9167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8530049171012755e-05,
+      "loss": 1.9933,
+      "step": 9168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8529736241783825e-05,
+      "loss": 2.014,
+      "step": 9169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8529423281892513e-05,
+      "loss": 1.7391,
+      "step": 9170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.852911029133994e-05,
+      "loss": 1.7712,
+      "step": 9171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8528797270127233e-05,
+      "loss": 2.0067,
+      "step": 9172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8528484218255515e-05,
+      "loss": 1.9711,
+      "step": 9173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8528171135725914e-05,
+      "loss": 1.9019,
+      "step": 9174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8527858022539558e-05,
+      "loss": 2.0301,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8527544878697562e-05,
+      "loss": 1.8368,
+      "step": 9176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8527231704201067e-05,
+      "loss": 1.8727,
+      "step": 9177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8526918499051185e-05,
+      "loss": 1.9888,
+      "step": 9178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8526605263249048e-05,
+      "loss": 1.9484,
+      "step": 9179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8526291996795787e-05,
+      "loss": 1.7876,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8525978699692516e-05,
+      "loss": 2.0877,
+      "step": 9181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8525665371940372e-05,
+      "loss": 1.8737,
+      "step": 9182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8525352013540475e-05,
+      "loss": 1.6554,
+      "step": 9183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8525038624493957e-05,
+      "loss": 1.9064,
+      "step": 9184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.852472520480194e-05,
+      "loss": 1.7372,
+      "step": 9185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8524411754465546e-05,
+      "loss": 1.8027,
+      "step": 9186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8524098273485914e-05,
+      "loss": 1.6808,
+      "step": 9187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8523784761864164e-05,
+      "loss": 1.7102,
+      "step": 9188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.852347121960142e-05,
+      "loss": 1.8184,
+      "step": 9189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8523157646698818e-05,
+      "loss": 1.7497,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8522844043157476e-05,
+      "loss": 2.0362,
+      "step": 9191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.852253040897853e-05,
+      "loss": 1.9198,
+      "step": 9192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8522216744163097e-05,
+      "loss": 1.9659,
+      "step": 9193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8521903048712314e-05,
+      "loss": 1.9711,
+      "step": 9194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.8521589322627303e-05,
+      "loss": 1.445,
+      "step": 9195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.852127556590919e-05,
+      "loss": 1.7694,
+      "step": 9196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8520961778559113e-05,
+      "loss": 2.1696,
+      "step": 9197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8520647960578194e-05,
+      "loss": 1.8775,
+      "step": 9198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8520334111967557e-05,
+      "loss": 2.0174,
+      "step": 9199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8520020232728333e-05,
+      "loss": 1.7676,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8519706322861656e-05,
+      "loss": 1.8185,
+      "step": 9201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8519392382368645e-05,
+      "loss": 1.4692,
+      "step": 9202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8519078411250434e-05,
+      "loss": 1.7821,
+      "step": 9203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.8518764409508152e-05,
+      "loss": 3.2777,
+      "step": 9204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8518450377142926e-05,
+      "loss": 1.8673,
+      "step": 9205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8518136314155887e-05,
+      "loss": 1.7768,
+      "step": 9206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.851782222054816e-05,
+      "loss": 1.6618,
+      "step": 9207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.851750809632088e-05,
+      "loss": 1.79,
+      "step": 9208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8517193941475166e-05,
+      "loss": 2.5863,
+      "step": 9209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.851687975601216e-05,
+      "loss": 2.0088,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8516565539932983e-05,
+      "loss": 1.9544,
+      "step": 9211
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8516251293238768e-05,
+      "loss": 1.9208,
+      "step": 9212
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.851593701593064e-05,
+      "loss": 2.0825,
+      "step": 9213
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.851562270800974e-05,
+      "loss": 2.0154,
+      "step": 9214
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8515308369477185e-05,
+      "loss": 1.7796,
+      "step": 9215
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8514994000334112e-05,
+      "loss": 1.7638,
+      "step": 9216
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625,
+      "learning_rate": 1.851467960058165e-05,
+      "loss": 1.6906,
+      "step": 9217
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.8514365170220925e-05,
+      "loss": 1.7759,
+      "step": 9218
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8514050709253073e-05,
+      "loss": 2.064,
+      "step": 9219
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8513736217679225e-05,
+      "loss": 2.145,
+      "step": 9220
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8513421695500503e-05,
+      "loss": 1.9293,
+      "step": 9221
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8513107142718048e-05,
+      "loss": 1.719,
+      "step": 9222
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.851279255933298e-05,
+      "loss": 2.0034,
+      "step": 9223
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8512477945346442e-05,
+      "loss": 1.8724,
+      "step": 9224
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.851216330075956e-05,
+      "loss": 1.8107,
+      "step": 9225
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.851184862557346e-05,
+      "loss": 1.9397,
+      "step": 9226
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8511533919789277e-05,
+      "loss": 2.1113,
+      "step": 9227
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8511219183408144e-05,
+      "loss": 2.2276,
+      "step": 9228
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.8510904416431192e-05,
+      "loss": 1.5186,
+      "step": 9229
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8510589618859553e-05,
+      "loss": 1.9729,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.851027479069435e-05,
+      "loss": 1.9702,
+      "step": 9231
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8509959931936726e-05,
+      "loss": 1.9453,
+      "step": 9232
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.850964504258781e-05,
+      "loss": 1.7281,
+      "step": 9233
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8509330122648732e-05,
+      "loss": 2.2483,
+      "step": 9234
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.850901517212062e-05,
+      "loss": 1.8835,
+      "step": 9235
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.850870019100462e-05,
+      "loss": 1.6565,
+      "step": 9236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8508385179301848e-05,
+      "loss": 2.0253,
+      "step": 9237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8508070137013442e-05,
+      "loss": 1.7806,
+      "step": 9238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.850775506414054e-05,
+      "loss": 1.8425,
+      "step": 9239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.850743996068427e-05,
+      "loss": 2.6562,
+      "step": 9240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8507124826645765e-05,
+      "loss": 1.9115,
+      "step": 9241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.850680966202616e-05,
+      "loss": 1.8921,
+      "step": 9242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8506494466826583e-05,
+      "loss": 1.5365,
+      "step": 9243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.850617924104817e-05,
+      "loss": 1.6292,
+      "step": 9244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8505863984692055e-05,
+      "loss": 1.9876,
+      "step": 9245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.850554869775937e-05,
+      "loss": 1.8964,
+      "step": 9246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8505233380251254e-05,
+      "loss": 2.1882,
+      "step": 9247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.850491803216883e-05,
+      "loss": 1.9101,
+      "step": 9248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.850460265351324e-05,
+      "loss": 1.6666,
+      "step": 9249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.850428724428561e-05,
+      "loss": 1.9985,
+      "step": 9250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.850397180448708e-05,
+      "loss": 1.6441,
+      "step": 9251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8503656334118787e-05,
+      "loss": 1.8938,
+      "step": 9252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8503340833181857e-05,
+      "loss": 1.9995,
+      "step": 9253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.850302530167743e-05,
+      "loss": 1.7526,
+      "step": 9254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8502709739606635e-05,
+      "loss": 2.0342,
+      "step": 9255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.850239414697061e-05,
+      "loss": 1.9856,
+      "step": 9256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.850207852377049e-05,
+      "loss": 1.9068,
+      "step": 9257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.850176287000741e-05,
+      "loss": 1.9704,
+      "step": 9258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.85014471856825e-05,
+      "loss": 1.9962,
+      "step": 9259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8501131470796897e-05,
+      "loss": 1.8272,
+      "step": 9260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.850081572535174e-05,
+      "loss": 1.7877,
+      "step": 9261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.850049994934816e-05,
+      "loss": 1.9417,
+      "step": 9262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.850018414278729e-05,
+      "loss": 1.5626,
+      "step": 9263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8499868305670272e-05,
+      "loss": 1.7861,
+      "step": 9264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8499552437998237e-05,
+      "loss": 1.6592,
+      "step": 9265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.849923653977232e-05,
+      "loss": 2.1322,
+      "step": 9266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.849892061099366e-05,
+      "loss": 1.8189,
+      "step": 9267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8498604651663385e-05,
+      "loss": 1.9727,
+      "step": 9268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.849828866178264e-05,
+      "loss": 2.1668,
+      "step": 9269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8497972641352556e-05,
+      "loss": 1.9746,
+      "step": 9270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.849765659037427e-05,
+      "loss": 1.8777,
+      "step": 9271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8497340508848923e-05,
+      "loss": 2.167,
+      "step": 9272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.849702439677764e-05,
+      "loss": 1.7512,
+      "step": 9273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8496708254161567e-05,
+      "loss": 1.7775,
+      "step": 9274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8496392081001838e-05,
+      "loss": 1.9277,
+      "step": 9275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8496075877299585e-05,
+      "loss": 2.1744,
+      "step": 9276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.849575964305595e-05,
+      "loss": 2.1966,
+      "step": 9277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.849544337827207e-05,
+      "loss": 1.696,
+      "step": 9278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8495127082949082e-05,
+      "loss": 1.9493,
+      "step": 9279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8494810757088116e-05,
+      "loss": 2.0389,
+      "step": 9280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8494494400690315e-05,
+      "loss": 2.0541,
+      "step": 9281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8494178013756818e-05,
+      "loss": 2.0784,
+      "step": 9282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.849386159628876e-05,
+      "loss": 1.8907,
+      "step": 9283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8493545148287278e-05,
+      "loss": 1.8667,
+      "step": 9284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8493228669753506e-05,
+      "loss": 1.5824,
+      "step": 9285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.849291216068859e-05,
+      "loss": 1.5838,
+      "step": 9286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8492595621093664e-05,
+      "loss": 2.0472,
+      "step": 9287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.849227905096986e-05,
+      "loss": 1.9946,
+      "step": 9288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8491962450318325e-05,
+      "loss": 2.0154,
+      "step": 9289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8491645819140194e-05,
+      "loss": 1.9617,
+      "step": 9290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.84913291574366e-05,
+      "loss": 1.4652,
+      "step": 9291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.849101246520869e-05,
+      "loss": 1.9228,
+      "step": 9292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8490695742457598e-05,
+      "loss": 1.9261,
+      "step": 9293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.849037898918446e-05,
+      "loss": 1.7634,
+      "step": 9294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.849006220539042e-05,
+      "loss": 1.7053,
+      "step": 9295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8489745391076616e-05,
+      "loss": 1.9241,
+      "step": 9296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.848942854624418e-05,
+      "loss": 1.6695,
+      "step": 9297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.848911167089426e-05,
+      "loss": 1.7581,
+      "step": 9298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.848879476502799e-05,
+      "loss": 1.8634,
+      "step": 9299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8488477828646513e-05,
+      "loss": 2.0469,
+      "step": 9300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8488160861750962e-05,
+      "loss": 2.0859,
+      "step": 9301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.848784386434248e-05,
+      "loss": 1.8541,
+      "step": 9302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8487526836422206e-05,
+      "loss": 2.0338,
+      "step": 9303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.848720977799128e-05,
+      "loss": 1.8923,
+      "step": 9304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.848689268905085e-05,
+      "loss": 1.8009,
+      "step": 9305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.848657556960204e-05,
+      "loss": 1.7915,
+      "step": 9306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8486258419646e-05,
+      "loss": 1.5242,
+      "step": 9307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.8485941239183867e-05,
+      "loss": 2.1634,
+      "step": 9308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8485624028216782e-05,
+      "loss": 1.6993,
+      "step": 9309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.8485306786745886e-05,
+      "loss": 2.2174,
+      "step": 9310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.848498951477232e-05,
+      "loss": 1.4058,
+      "step": 9311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.848467221229722e-05,
+      "loss": 1.8783,
+      "step": 9312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8484354879321734e-05,
+      "loss": 1.4949,
+      "step": 9313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8484037515846998e-05,
+      "loss": 2.0104,
+      "step": 9314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8483720121874152e-05,
+      "loss": 1.9059,
+      "step": 9315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.848340269740434e-05,
+      "loss": 1.5091,
+      "step": 9316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.84830852424387e-05,
+      "loss": 1.649,
+      "step": 9317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8482767756978375e-05,
+      "loss": 1.8471,
+      "step": 9318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8482450241024507e-05,
+      "loss": 1.6963,
+      "step": 9319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8482132694578233e-05,
+      "loss": 2.1317,
+      "step": 9320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8481815117640704e-05,
+      "loss": 1.8595,
+      "step": 9321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.848149751021305e-05,
+      "loss": 1.9844,
+      "step": 9322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.848117987229642e-05,
+      "loss": 2.2608,
+      "step": 9323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8480862203891958e-05,
+      "loss": 1.9061,
+      "step": 9324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8480544505000793e-05,
+      "loss": 1.7285,
+      "step": 9325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8480226775624083e-05,
+      "loss": 2.1073,
+      "step": 9326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.847990901576296e-05,
+      "loss": 1.7342,
+      "step": 9327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8479591225418575e-05,
+      "loss": 1.6995,
+      "step": 9328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.847927340459206e-05,
+      "loss": 1.6608,
+      "step": 9329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8478955553284565e-05,
+      "loss": 1.7497,
+      "step": 9330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8478637671497228e-05,
+      "loss": 2.462,
+      "step": 9331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8478319759231197e-05,
+      "loss": 1.8019,
+      "step": 9332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.847800181648761e-05,
+      "loss": 1.9473,
+      "step": 9333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.847768384326761e-05,
+      "loss": 2.2383,
+      "step": 9334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.847736583957234e-05,
+      "loss": 2.1117,
+      "step": 9335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8477047805402946e-05,
+      "loss": 1.9215,
+      "step": 9336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.375,
+      "learning_rate": 1.847672974076057e-05,
+      "loss": 1.8823,
+      "step": 9337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8476411645646356e-05,
+      "loss": 1.7911,
+      "step": 9338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.847609352006145e-05,
+      "loss": 2.0593,
+      "step": 9339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.847577536400699e-05,
+      "loss": 2.0433,
+      "step": 9340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8475457177484118e-05,
+      "loss": 1.9654,
+      "step": 9341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8475138960493986e-05,
+      "loss": 1.9163,
+      "step": 9342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8474820713037732e-05,
+      "loss": 1.6025,
+      "step": 9343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8474502435116505e-05,
+      "loss": 1.7135,
+      "step": 9344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.847418412673144e-05,
+      "loss": 1.9639,
+      "step": 9345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.847386578788369e-05,
+      "loss": 1.8163,
+      "step": 9346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.84735474185744e-05,
+      "loss": 1.9573,
+      "step": 9347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8473229018804705e-05,
+      "loss": 1.8748,
+      "step": 9348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.847291058857576e-05,
+      "loss": 1.8707,
+      "step": 9349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8472592127888705e-05,
+      "loss": 2.0473,
+      "step": 9350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.847227363674468e-05,
+      "loss": 1.6552,
+      "step": 9351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.847195511514484e-05,
+      "loss": 1.7173,
+      "step": 9352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8471636563090322e-05,
+      "loss": 1.5326,
+      "step": 9353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8471317980582278e-05,
+      "loss": 1.9048,
+      "step": 9354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8470999367621847e-05,
+      "loss": 1.854,
+      "step": 9355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8470680724210176e-05,
+      "loss": 1.5315,
+      "step": 9356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8470362050348412e-05,
+      "loss": 1.9348,
+      "step": 9357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.8470043346037698e-05,
+      "loss": 2.1903,
+      "step": 9358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.846972461127918e-05,
+      "loss": 1.8429,
+      "step": 9359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8469405846074008e-05,
+      "loss": 2.0264,
+      "step": 9360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.8469087050423326e-05,
+      "loss": 1.9202,
+      "step": 9361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8468768224328278e-05,
+      "loss": 1.9812,
+      "step": 9362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.846844936779001e-05,
+      "loss": 1.8874,
+      "step": 9363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.846813048080967e-05,
+      "loss": 1.9645,
+      "step": 9364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8467811563388403e-05,
+      "loss": 2.015,
+      "step": 9365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8467492615527357e-05,
+      "loss": 1.7982,
+      "step": 9366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8467173637227675e-05,
+      "loss": 2.0243,
+      "step": 9367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.846685462849051e-05,
+      "loss": 1.8942,
+      "step": 9368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8466535589317e-05,
+      "loss": 1.9586,
+      "step": 9369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8466216519708303e-05,
+      "loss": 1.8167,
+      "step": 9370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8465897419665558e-05,
+      "loss": 1.708,
+      "step": 9371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.846557828918991e-05,
+      "loss": 2.2172,
+      "step": 9372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8465259128282512e-05,
+      "loss": 1.723,
+      "step": 9373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.846493993694451e-05,
+      "loss": 1.6111,
+      "step": 9374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.846462071517705e-05,
+      "loss": 1.9166,
+      "step": 9375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8464301462981282e-05,
+      "loss": 1.8602,
+      "step": 9376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.846398218035835e-05,
+      "loss": 2.0859,
+      "step": 9377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8463662867309405e-05,
+      "loss": 1.9336,
+      "step": 9378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8463343523835594e-05,
+      "loss": 2.1929,
+      "step": 9379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8463024149938064e-05,
+      "loss": 1.683,
+      "step": 9380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8462704745617963e-05,
+      "loss": 1.6898,
+      "step": 9381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8462385310876444e-05,
+      "loss": 2.17,
+      "step": 9382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8462065845714644e-05,
+      "loss": 1.9232,
+      "step": 9383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8461746350133722e-05,
+      "loss": 1.6297,
+      "step": 9384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8461426824134825e-05,
+      "loss": 1.6232,
+      "step": 9385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.84611072677191e-05,
+      "loss": 1.7546,
+      "step": 9386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.375,
+      "learning_rate": 1.846078768088769e-05,
+      "loss": 2.0923,
+      "step": 9387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8460468063641756e-05,
+      "loss": 1.789,
+      "step": 9388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.846014841598243e-05,
+      "loss": 2.0764,
+      "step": 9389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.845982873791088e-05,
+      "loss": 1.8032,
+      "step": 9390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5,
+      "learning_rate": 1.8459509029428245e-05,
+      "loss": 1.8368,
+      "step": 9391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8459189290535676e-05,
+      "loss": 1.9539,
+      "step": 9392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.845886952123432e-05,
+      "loss": 1.6525,
+      "step": 9393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8458549721525328e-05,
+      "loss": 1.9605,
+      "step": 9394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8458229891409852e-05,
+      "loss": 1.8905,
+      "step": 9395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.845791003088904e-05,
+      "loss": 1.6357,
+      "step": 9396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.845759013996404e-05,
+      "loss": 2.0777,
+      "step": 9397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8457270218636004e-05,
+      "loss": 1.9236,
+      "step": 9398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8456950266906085e-05,
+      "loss": 1.8776,
+      "step": 9399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.845663028477543e-05,
+      "loss": 1.8641,
+      "step": 9400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.845631027224518e-05,
+      "loss": 1.8165,
+      "step": 9401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.84559902293165e-05,
+      "loss": 1.6924,
+      "step": 9402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8455670155990538e-05,
+      "loss": 1.9056,
+      "step": 9403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.845535005226844e-05,
+      "loss": 1.7858,
+      "step": 9404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8455029918151356e-05,
+      "loss": 1.5489,
+      "step": 9405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.8454709753640442e-05,
+      "loss": 1.7963,
+      "step": 9406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.845438955873684e-05,
+      "loss": 1.8734,
+      "step": 9407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8454069333441715e-05,
+      "loss": 1.7903,
+      "step": 9408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8453749077756205e-05,
+      "loss": 1.8471,
+      "step": 9409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8453428791681467e-05,
+      "loss": 1.7053,
+      "step": 9410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8453108475218653e-05,
+      "loss": 1.9545,
+      "step": 9411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.845278812836891e-05,
+      "loss": 2.0936,
+      "step": 9412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8452467751133395e-05,
+      "loss": 1.9216,
+      "step": 9413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8452147343513257e-05,
+      "loss": 1.8165,
+      "step": 9414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.845182690550965e-05,
+      "loss": 2.2786,
+      "step": 9415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8451506437123723e-05,
+      "loss": 1.8794,
+      "step": 9416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8451185938356627e-05,
+      "loss": 1.8105,
+      "step": 9417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.845086540920952e-05,
+      "loss": 1.5961,
+      "step": 9418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8450544849683546e-05,
+      "loss": 1.7179,
+      "step": 9419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8450224259779866e-05,
+      "loss": 1.7684,
+      "step": 9420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8449903639499625e-05,
+      "loss": 1.6061,
+      "step": 9421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.844958298884398e-05,
+      "loss": 1.5531,
+      "step": 9422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8449262307814083e-05,
+      "loss": 1.9487,
+      "step": 9423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8448941596411085e-05,
+      "loss": 1.979,
+      "step": 9424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.844862085463614e-05,
+      "loss": 1.9687,
+      "step": 9425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8448300082490402e-05,
+      "loss": 1.7537,
+      "step": 9426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8447979279975024e-05,
+      "loss": 2.1233,
+      "step": 9427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.8447658447091157e-05,
+      "loss": 2.0852,
+      "step": 9428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8447337583839956e-05,
+      "loss": 1.8802,
+      "step": 9429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8447016690222573e-05,
+      "loss": 1.7097,
+      "step": 9430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8446695766240164e-05,
+      "loss": 1.9229,
+      "step": 9431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8446374811893883e-05,
+      "loss": 1.5879,
+      "step": 9432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.844605382718488e-05,
+      "loss": 1.9314,
+      "step": 9433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.844573281211431e-05,
+      "loss": 2.0367,
+      "step": 9434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.844541176668333e-05,
+      "loss": 1.8036,
+      "step": 9435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.844509069089309e-05,
+      "loss": 2.0074,
+      "step": 9436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.844476958474475e-05,
+      "loss": 2.1455,
+      "step": 9437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8444448448239455e-05,
+      "loss": 1.7602,
+      "step": 9438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8444127281378367e-05,
+      "loss": 1.9064,
+      "step": 9439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8443806084162638e-05,
+      "loss": 1.6138,
+      "step": 9440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8443484856593425e-05,
+      "loss": 1.8311,
+      "step": 9441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.844316359867188e-05,
+      "loss": 1.8684,
+      "step": 9442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8442842310399158e-05,
+      "loss": 1.8521,
+      "step": 9443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8442520991776417e-05,
+      "loss": 1.9035,
+      "step": 9444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8442199642804805e-05,
+      "loss": 2.0409,
+      "step": 9445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8441878263485485e-05,
+      "loss": 1.4754,
+      "step": 9446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.844155685381961e-05,
+      "loss": 1.9915,
+      "step": 9447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8441235413808332e-05,
+      "loss": 2.1609,
+      "step": 9448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.844091394345281e-05,
+      "loss": 1.8069,
+      "step": 9449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8440592442754197e-05,
+      "loss": 1.617,
+      "step": 9450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8440270911713655e-05,
+      "loss": 2.1853,
+      "step": 9451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8439949350332332e-05,
+      "loss": 1.7502,
+      "step": 9452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8439627758611384e-05,
+      "loss": 1.7636,
+      "step": 9453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.843930613655197e-05,
+      "loss": 1.9061,
+      "step": 9454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8438984484155254e-05,
+      "loss": 2.0099,
+      "step": 9455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8438662801422377e-05,
+      "loss": 1.525,
+      "step": 9456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8438341088354506e-05,
+      "loss": 1.6055,
+      "step": 9457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8438019344952796e-05,
+      "loss": 1.6265,
+      "step": 9458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.84376975712184e-05,
+      "loss": 1.9396,
+      "step": 9459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8437375767152478e-05,
+      "loss": 1.9023,
+      "step": 9460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8437053932756183e-05,
+      "loss": 2.0176,
+      "step": 9461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8436732068030678e-05,
+      "loss": 1.9082,
+      "step": 9462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8436410172977114e-05,
+      "loss": 1.8232,
+      "step": 9463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.843608824759665e-05,
+      "loss": 2.0782,
+      "step": 9464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8435766291890445e-05,
+      "loss": 1.9242,
+      "step": 9465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8435444305859654e-05,
+      "loss": 1.7541,
+      "step": 9466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.8435122289505435e-05,
+      "loss": 2.0965,
+      "step": 9467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.843480024282895e-05,
+      "loss": 1.8269,
+      "step": 9468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8434478165831352e-05,
+      "loss": 1.9256,
+      "step": 9469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.84341560585138e-05,
+      "loss": 1.9035,
+      "step": 9470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.843383392087745e-05,
+      "loss": 1.884,
+      "step": 9471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.843351175292346e-05,
+      "loss": 1.8538,
+      "step": 9472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.843318955465299e-05,
+      "loss": 1.7819,
+      "step": 9473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8432867326067197e-05,
+      "loss": 1.9568,
+      "step": 9474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8432545067167246e-05,
+      "loss": 1.8977,
+      "step": 9475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8432222777954286e-05,
+      "loss": 2.0147,
+      "step": 9476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.843190045842948e-05,
+      "loss": 1.4913,
+      "step": 9477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8431578108593982e-05,
+      "loss": 2.0605,
+      "step": 9478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.843125572844896e-05,
+      "loss": 2.2352,
+      "step": 9479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8430933317995566e-05,
+      "loss": 1.7832,
+      "step": 9480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.843061087723496e-05,
+      "loss": 1.5304,
+      "step": 9481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8430288406168303e-05,
+      "loss": 1.9743,
+      "step": 9482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.842996590479675e-05,
+      "loss": 1.6514,
+      "step": 9483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8429643373121467e-05,
+      "loss": 2.1294,
+      "step": 9484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8429320811143604e-05,
+      "loss": 1.8633,
+      "step": 9485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8428998218864332e-05,
+      "loss": 2.0885,
+      "step": 9486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8428675596284805e-05,
+      "loss": 1.613,
+      "step": 9487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8428352943406177e-05,
+      "loss": 1.9301,
+      "step": 9488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.842803026022962e-05,
+      "loss": 2.3131,
+      "step": 9489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.842770754675628e-05,
+      "loss": 2.2873,
+      "step": 9490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.842738480298733e-05,
+      "loss": 1.8955,
+      "step": 9491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8427062028923927e-05,
+      "loss": 1.8157,
+      "step": 9492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8426739224567222e-05,
+      "loss": 1.7611,
+      "step": 9493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8426416389918387e-05,
+      "loss": 2.0159,
+      "step": 9494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8426093524978577e-05,
+      "loss": 1.9243,
+      "step": 9495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8425770629748952e-05,
+      "loss": 1.6801,
+      "step": 9496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8425447704230673e-05,
+      "loss": 1.9131,
+      "step": 9497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8425124748424905e-05,
+      "loss": 2.0164,
+      "step": 9498
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8424801762332806e-05,
+      "loss": 1.9169,
+      "step": 9499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8424478745955533e-05,
+      "loss": 2.0416,
+      "step": 9500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8424155699294256e-05,
+      "loss": 2.0032,
+      "step": 9501
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8423832622350132e-05,
+      "loss": 1.8021,
+      "step": 9502
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.842350951512432e-05,
+      "loss": 1.4412,
+      "step": 9503
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8423186377617983e-05,
+      "loss": 2.0887,
+      "step": 9504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8422863209832282e-05,
+      "loss": 1.9099,
+      "step": 9505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8422540011768382e-05,
+      "loss": 2.3006,
+      "step": 9506
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.842221678342744e-05,
+      "loss": 2.1688,
+      "step": 9507
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.842189352481062e-05,
+      "loss": 1.7246,
+      "step": 9508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8421570235919087e-05,
+      "loss": 1.7659,
+      "step": 9509
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8421246916754003e-05,
+      "loss": 1.549,
+      "step": 9510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8420923567316523e-05,
+      "loss": 1.9714,
+      "step": 9511
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.8420600187607818e-05,
+      "loss": 2.0004,
+      "step": 9512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8420276777629044e-05,
+      "loss": 1.6502,
+      "step": 9513
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8419953337381366e-05,
+      "loss": 1.6986,
+      "step": 9514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8419629866865948e-05,
+      "loss": 1.9985,
+      "step": 9515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8419306366083955e-05,
+      "loss": 1.8984,
+      "step": 9516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8418982835036543e-05,
+      "loss": 2.0093,
+      "step": 9517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8418659273724882e-05,
+      "loss": 1.7618,
+      "step": 9518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.841833568215013e-05,
+      "loss": 1.937,
+      "step": 9519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.8418012060313447e-05,
+      "loss": 2.2462,
+      "step": 9520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8417688408216008e-05,
+      "loss": 1.9446,
+      "step": 9521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8417364725858968e-05,
+      "loss": 1.8736,
+      "step": 9522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8417041013243494e-05,
+      "loss": 1.9172,
+      "step": 9523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8416717270370744e-05,
+      "loss": 1.7124,
+      "step": 9524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8416393497241892e-05,
+      "loss": 1.9466,
+      "step": 9525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.841606969385809e-05,
+      "loss": 2.0185,
+      "step": 9526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8415745860220512e-05,
+      "loss": 1.7652,
+      "step": 9527
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8415421996330315e-05,
+      "loss": 1.8987,
+      "step": 9528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8415098102188667e-05,
+      "loss": 1.4861,
+      "step": 9529
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.841477417779673e-05,
+      "loss": 2.0529,
+      "step": 9530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.841445022315567e-05,
+      "loss": 1.6837,
+      "step": 9531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8414126238266656e-05,
+      "loss": 1.9824,
+      "step": 9532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8413802223130842e-05,
+      "loss": 1.9387,
+      "step": 9533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8413478177749402e-05,
+      "loss": 1.8007,
+      "step": 9534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.84131541021235e-05,
+      "loss": 1.5721,
+      "step": 9535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8412829996254295e-05,
+      "loss": 2.0241,
+      "step": 9536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8412505860142955e-05,
+      "loss": 1.7108,
+      "step": 9537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.841218169379065e-05,
+      "loss": 1.8688,
+      "step": 9538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8411857497198536e-05,
+      "loss": 1.8631,
+      "step": 9539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.841153327036779e-05,
+      "loss": 1.619,
+      "step": 9540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8411209013299568e-05,
+      "loss": 1.7124,
+      "step": 9541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.841088472599504e-05,
+      "loss": 0.0,
+      "step": 9542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8410560408455364e-05,
+      "loss": 1.7385,
+      "step": 9543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.841023606068172e-05,
+      "loss": 1.8388,
+      "step": 9544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8409911682675262e-05,
+      "loss": 2.0997,
+      "step": 9545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8409587274437162e-05,
+      "loss": 1.8029,
+      "step": 9546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8409262835968586e-05,
+      "loss": 1.7533,
+      "step": 9547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.8408938367270697e-05,
+      "loss": 1.6505,
+      "step": 9548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8408613868344663e-05,
+      "loss": 2.0178,
+      "step": 9549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.840828933919165e-05,
+      "loss": 2.061,
+      "step": 9550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8407964779812828e-05,
+      "loss": 1.9188,
+      "step": 9551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8407640190209362e-05,
+      "loss": 2.1151,
+      "step": 9552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8407315570382412e-05,
+      "loss": 1.7323,
+      "step": 9553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8406990920333157e-05,
+      "loss": 1.7755,
+      "step": 9554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8406666240062756e-05,
+      "loss": 2.0424,
+      "step": 9555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8406341529572377e-05,
+      "loss": 1.9458,
+      "step": 9556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8406016788863187e-05,
+      "loss": 2.6607,
+      "step": 9557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.840569201793636e-05,
+      "loss": 1.9845,
+      "step": 9558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.840536721679305e-05,
+      "loss": 1.5855,
+      "step": 9559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.840504238543444e-05,
+      "loss": 2.0913,
+      "step": 9560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8404717523861685e-05,
+      "loss": 1.892,
+      "step": 9561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8404392632075963e-05,
+      "loss": 2.0499,
+      "step": 9562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.840406771007843e-05,
+      "loss": 2.1589,
+      "step": 9563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8403742757870268e-05,
+      "loss": 1.7999,
+      "step": 9564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.840341777545264e-05,
+      "loss": 1.7413,
+      "step": 9565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8403092762826707e-05,
+      "loss": 1.8959,
+      "step": 9566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8402767719993644e-05,
+      "loss": 1.7669,
+      "step": 9567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8402442646954617e-05,
+      "loss": 1.812,
+      "step": 9568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8402117543710795e-05,
+      "loss": 1.9245,
+      "step": 9569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.840179241026335e-05,
+      "loss": 2.0315,
+      "step": 9570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.840146724661345e-05,
+      "loss": 1.6484,
+      "step": 9571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.840114205276226e-05,
+      "loss": 1.9102,
+      "step": 9572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.840081682871095e-05,
+      "loss": 1.7385,
+      "step": 9573
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8400491574460692e-05,
+      "loss": 2.0413,
+      "step": 9574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8400166290012654e-05,
+      "loss": 1.8157,
+      "step": 9575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8399840975368002e-05,
+      "loss": 2.0916,
+      "step": 9576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8399515630527907e-05,
+      "loss": 1.8254,
+      "step": 9577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8399190255493544e-05,
+      "loss": 2.0915,
+      "step": 9578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8398864850266078e-05,
+      "loss": 1.9869,
+      "step": 9579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8398539414846678e-05,
+      "loss": 1.8481,
+      "step": 9580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8398213949236517e-05,
+      "loss": 1.9021,
+      "step": 9581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.839788845343676e-05,
+      "loss": 2.0734,
+      "step": 9582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.8397562927448582e-05,
+      "loss": 1.8804,
+      "step": 9583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.839723737127315e-05,
+      "loss": 2.1004,
+      "step": 9584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8396911784911636e-05,
+      "loss": 1.8432,
+      "step": 9585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8396586168365208e-05,
+      "loss": 1.8209,
+      "step": 9586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.839626052163504e-05,
+      "loss": 2.0758,
+      "step": 9587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8395934844722303e-05,
+      "loss": 1.9103,
+      "step": 9588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8395609137628165e-05,
+      "loss": 1.6076,
+      "step": 9589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.8395283400353795e-05,
+      "loss": 2.0581,
+      "step": 9590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.839495763290037e-05,
+      "loss": 1.7626,
+      "step": 9591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8394631835269055e-05,
+      "loss": 1.551,
+      "step": 9592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8394306007461026e-05,
+      "loss": 1.9793,
+      "step": 9593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.839398014947745e-05,
+      "loss": 1.7605,
+      "step": 9594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8393654261319504e-05,
+      "loss": 2.1549,
+      "step": 9595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.839332834298835e-05,
+      "loss": 2.038,
+      "step": 9596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.839300239448517e-05,
+      "loss": 1.9623,
+      "step": 9597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8392676415811132e-05,
+      "loss": 2.0623,
+      "step": 9598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.83923504069674e-05,
+      "loss": 1.8447,
+      "step": 9599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8392024367955163e-05,
+      "loss": 2.0215,
+      "step": 9600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8391698298775578e-05,
+      "loss": 1.8895,
+      "step": 9601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8391372199429823e-05,
+      "loss": 1.9211,
+      "step": 9602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.839104606991907e-05,
+      "loss": 2.0614,
+      "step": 9603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8390719910244487e-05,
+      "loss": 1.7461,
+      "step": 9604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8390393720407255e-05,
+      "loss": 1.9248,
+      "step": 9605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.839006750040854e-05,
+      "loss": 2.012,
+      "step": 9606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8389741250249518e-05,
+      "loss": 1.8407,
+      "step": 9607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.838941496993136e-05,
+      "loss": 1.8168,
+      "step": 9608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8389088659455235e-05,
+      "loss": 1.9187,
+      "step": 9609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8388762318822326e-05,
+      "loss": 1.9323,
+      "step": 9610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8388435948033794e-05,
+      "loss": 1.8015,
+      "step": 9611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.838810954709082e-05,
+      "loss": 1.5661,
+      "step": 9612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8387783115994583e-05,
+      "loss": 1.9713,
+      "step": 9613
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8387456654746242e-05,
+      "loss": 1.717,
+      "step": 9614
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.838713016334698e-05,
+      "loss": 1.96,
+      "step": 9615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8386803641797972e-05,
+      "loss": 1.5911,
+      "step": 9616
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8386477090100383e-05,
+      "loss": 1.7737,
+      "step": 9617
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8386150508255395e-05,
+      "loss": 2.0953,
+      "step": 9618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8385823896264178e-05,
+      "loss": 2.1939,
+      "step": 9619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.838549725412791e-05,
+      "loss": 2.0214,
+      "step": 9620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.838517058184776e-05,
+      "loss": 1.8416,
+      "step": 9621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8384843879424905e-05,
+      "loss": 2.0424,
+      "step": 9622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.838451714686052e-05,
+      "loss": 1.84,
+      "step": 9623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.838419038415578e-05,
+      "loss": 2.1512,
+      "step": 9624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8383863591311855e-05,
+      "loss": 2.1728,
+      "step": 9625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8383536768329925e-05,
+      "loss": 1.7747,
+      "step": 9626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8383209915211163e-05,
+      "loss": 1.4828,
+      "step": 9627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8382883031956744e-05,
+      "loss": 1.8219,
+      "step": 9628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8382556118567844e-05,
+      "loss": 1.863,
+      "step": 9629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8382229175045634e-05,
+      "loss": 2.0801,
+      "step": 9630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8381902201391292e-05,
+      "loss": 1.9395,
+      "step": 9631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8381575197606e-05,
+      "loss": 2.2181,
+      "step": 9632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.838124816369092e-05,
+      "loss": 1.6324,
+      "step": 9633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.838092109964724e-05,
+      "loss": 1.7808,
+      "step": 9634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8380594005476127e-05,
+      "loss": 2.1083,
+      "step": 9635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8380266881178764e-05,
+      "loss": 1.9597,
+      "step": 9636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8379939726756322e-05,
+      "loss": 1.5441,
+      "step": 9637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.837961254220998e-05,
+      "loss": 1.8625,
+      "step": 9638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.837928532754091e-05,
+      "loss": 1.9233,
+      "step": 9639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8378958082750294e-05,
+      "loss": 1.8481,
+      "step": 9640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8378630807839302e-05,
+      "loss": 1.8522,
+      "step": 9641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8378303502809113e-05,
+      "loss": 2.0369,
+      "step": 9642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.837797616766091e-05,
+      "loss": 1.7085,
+      "step": 9643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.837764880239586e-05,
+      "loss": 2.4724,
+      "step": 9644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8377321407015142e-05,
+      "loss": 1.7418,
+      "step": 9645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8376993981519937e-05,
+      "loss": 1.9245,
+      "step": 9646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.837666652591142e-05,
+      "loss": 1.583,
+      "step": 9647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.837633904019077e-05,
+      "loss": 1.9715,
+      "step": 9648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.837601152435916e-05,
+      "loss": 1.8415,
+      "step": 9649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8375683978417767e-05,
+      "loss": 2.5407,
+      "step": 9650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8375356402367775e-05,
+      "loss": 1.8787,
+      "step": 9651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8375028796210353e-05,
+      "loss": 2.0588,
+      "step": 9652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8374701159946683e-05,
+      "loss": 1.8163,
+      "step": 9653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8374373493577948e-05,
+      "loss": 2.0216,
+      "step": 9654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.837404579710532e-05,
+      "loss": 1.8944,
+      "step": 9655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8373718070529976e-05,
+      "loss": 1.9696,
+      "step": 9656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8373390313853095e-05,
+      "loss": 2.2053,
+      "step": 9657
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8373062527075855e-05,
+      "loss": 1.7438,
+      "step": 9658
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8372734710199438e-05,
+      "loss": 1.6687,
+      "step": 9659
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8372406863225022e-05,
+      "loss": 2.0532,
+      "step": 9660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8372078986153775e-05,
+      "loss": 1.8842,
+      "step": 9661
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.837175107898689e-05,
+      "loss": 2.4314,
+      "step": 9662
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.837142314172554e-05,
+      "loss": 2.0037,
+      "step": 9663
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8371095174370905e-05,
+      "loss": 1.8849,
+      "step": 9664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.837076717692416e-05,
+      "loss": 1.6429,
+      "step": 9665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8370439149386484e-05,
+      "loss": 2.0981,
+      "step": 9666
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8370111091759065e-05,
+      "loss": 2.1005,
+      "step": 9667
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8369783004043073e-05,
+      "loss": 1.5503,
+      "step": 9668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8369454886239692e-05,
+      "loss": 1.8306,
+      "step": 9669
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8369126738350097e-05,
+      "loss": 1.8601,
+      "step": 9670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.836879856037547e-05,
+      "loss": 2.1753,
+      "step": 9671
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8368470352316995e-05,
+      "loss": 1.9003,
+      "step": 9672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8368142114175843e-05,
+      "loss": 1.7579,
+      "step": 9673
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8367813845953203e-05,
+      "loss": 1.9021,
+      "step": 9674
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8367485547650252e-05,
+      "loss": 1.6156,
+      "step": 9675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.836715721926817e-05,
+      "loss": 1.7976,
+      "step": 9676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.8366828860808134e-05,
+      "loss": 1.956,
+      "step": 9677
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.8366500472271325e-05,
+      "loss": 2.1998,
+      "step": 9678
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.836617205365893e-05,
+      "loss": 1.8285,
+      "step": 9679
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8365843604972124e-05,
+      "loss": 2.1328,
+      "step": 9680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8365515126212087e-05,
+      "loss": 1.938,
+      "step": 9681
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.836518661738e-05,
+      "loss": 1.6181,
+      "step": 9682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.836485807847705e-05,
+      "loss": 1.835,
+      "step": 9683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8364529509504412e-05,
+      "loss": 1.9164,
+      "step": 9684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8364200910463266e-05,
+      "loss": 1.6838,
+      "step": 9685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8363872281354796e-05,
+      "loss": 1.7171,
+      "step": 9686
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8363543622180186e-05,
+      "loss": 2.0971,
+      "step": 9687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8363214932940613e-05,
+      "loss": 1.9773,
+      "step": 9688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8362886213637263e-05,
+      "loss": 2.0877,
+      "step": 9689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.836255746427131e-05,
+      "loss": 1.75,
+      "step": 9690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8362228684843944e-05,
+      "loss": 1.5979,
+      "step": 9691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.836189987535634e-05,
+      "loss": 1.9459,
+      "step": 9692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8361571035809688e-05,
+      "loss": 1.9154,
+      "step": 9693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8361242166205162e-05,
+      "loss": 1.7079,
+      "step": 9694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.836091326654395e-05,
+      "loss": 2.0756,
+      "step": 9695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8360584336827234e-05,
+      "loss": 1.9883,
+      "step": 9696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.836025537705619e-05,
+      "loss": 1.9255,
+      "step": 9697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8359926387232006e-05,
+      "loss": 1.8553,
+      "step": 9698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8359597367355863e-05,
+      "loss": 1.8512,
+      "step": 9699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.835926831742895e-05,
+      "loss": 1.6077,
+      "step": 9700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8358939237452437e-05,
+      "loss": 1.9324,
+      "step": 9701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.835861012742752e-05,
+      "loss": 2.0983,
+      "step": 9702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.835828098735537e-05,
+      "loss": 2.0954,
+      "step": 9703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8357951817237183e-05,
+      "loss": 1.8975,
+      "step": 9704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8357622617074133e-05,
+      "loss": 1.7701,
+      "step": 9705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8357293386867407e-05,
+      "loss": 1.7477,
+      "step": 9706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8356964126618187e-05,
+      "loss": 1.8843,
+      "step": 9707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.835663483632766e-05,
+      "loss": 1.738,
+      "step": 9708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8356305515997e-05,
+      "loss": 1.9032,
+      "step": 9709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8355976165627406e-05,
+      "loss": 1.6724,
+      "step": 9710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8355646785220047e-05,
+      "loss": 1.7098,
+      "step": 9711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8355317374776118e-05,
+      "loss": 1.8774,
+      "step": 9712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8354987934296796e-05,
+      "loss": 1.6516,
+      "step": 9713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8354658463783272e-05,
+      "loss": 1.9343,
+      "step": 9714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8354328963236728e-05,
+      "loss": 2.0768,
+      "step": 9715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8353999432658342e-05,
+      "loss": 1.9559,
+      "step": 9716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8353669872049305e-05,
+      "loss": 1.8647,
+      "step": 9717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.83533402814108e-05,
+      "loss": 1.5995,
+      "step": 9718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8353010660744018e-05,
+      "loss": 1.8173,
+      "step": 9719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.835268101005013e-05,
+      "loss": 2.1535,
+      "step": 9720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.8352351329330337e-05,
+      "loss": 2.1978,
+      "step": 9721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.835202161858581e-05,
+      "loss": 1.8869,
+      "step": 9722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8351691877817744e-05,
+      "loss": 1.7482,
+      "step": 9723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8351362107027317e-05,
+      "loss": 1.6643,
+      "step": 9724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8351032306215718e-05,
+      "loss": 1.7375,
+      "step": 9725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.835070247538414e-05,
+      "loss": 1.5403,
+      "step": 9726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8350372614533753e-05,
+      "loss": 1.7372,
+      "step": 9727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8350042723665757e-05,
+      "loss": 1.9094,
+      "step": 9728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.834971280278133e-05,
+      "loss": 1.9025,
+      "step": 9729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.834938285188166e-05,
+      "loss": 2.0009,
+      "step": 9730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8349052870967932e-05,
+      "loss": 1.998,
+      "step": 9731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8348722860041337e-05,
+      "loss": 1.8091,
+      "step": 9732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8348392819103056e-05,
+      "loss": 2.0186,
+      "step": 9733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.834806274815428e-05,
+      "loss": 1.9981,
+      "step": 9734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8347732647196186e-05,
+      "loss": 1.7881,
+      "step": 9735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.834740251622997e-05,
+      "loss": 1.8829,
+      "step": 9736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.834707235525682e-05,
+      "loss": 2.0828,
+      "step": 9737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8346742164277917e-05,
+      "loss": 1.7859,
+      "step": 9738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.8346411943294448e-05,
+      "loss": 2.136,
+      "step": 9739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.83460816923076e-05,
+      "loss": 1.7457,
+      "step": 9740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8345751411318568e-05,
+      "loss": 1.8717,
+      "step": 9741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8345421100328528e-05,
+      "loss": 1.8437,
+      "step": 9742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8345090759338675e-05,
+      "loss": 2.2726,
+      "step": 9743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8344760388350195e-05,
+      "loss": 1.6065,
+      "step": 9744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8344429987364273e-05,
+      "loss": 1.8766,
+      "step": 9745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.83440995563821e-05,
+      "loss": 1.9121,
+      "step": 9746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8343769095404862e-05,
+      "loss": 1.759,
+      "step": 9747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.8343438604433748e-05,
+      "loss": 2.3196,
+      "step": 9748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8343108083469946e-05,
+      "loss": 1.8141,
+      "step": 9749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.8342777532514642e-05,
+      "loss": 1.9076,
+      "step": 9750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8342446951569025e-05,
+      "loss": 1.5835,
+      "step": 9751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.8342116340634285e-05,
+      "loss": 2.0515,
+      "step": 9752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.834178569971161e-05,
+      "loss": 1.6582,
+      "step": 9753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8341455028802187e-05,
+      "loss": 1.8409,
+      "step": 9754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8341124327907208e-05,
+      "loss": 1.7133,
+      "step": 9755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.834079359702786e-05,
+      "loss": 1.8105,
+      "step": 9756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.834046283616533e-05,
+      "loss": 1.9965,
+      "step": 9757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8340132045320808e-05,
+      "loss": 1.6759,
+      "step": 9758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8339801224495484e-05,
+      "loss": 1.8647,
+      "step": 9759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8339470373690547e-05,
+      "loss": 1.8869,
+      "step": 9760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.8339139492907187e-05,
+      "loss": 2.1827,
+      "step": 9761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.833880858214659e-05,
+      "loss": 2.0323,
+      "step": 9762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8338477641409954e-05,
+      "loss": 1.9676,
+      "step": 9763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8338146670698457e-05,
+      "loss": 2.0034,
+      "step": 9764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8337815670013297e-05,
+      "loss": 2.04,
+      "step": 9765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.8337484639355662e-05,
+      "loss": 1.7716,
+      "step": 9766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.833715357872674e-05,
+      "loss": 1.6751,
+      "step": 9767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8336822488127727e-05,
+      "loss": 1.8563,
+      "step": 9768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8336491367559802e-05,
+      "loss": 1.9877,
+      "step": 9769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8336160217024166e-05,
+      "loss": 1.9075,
+      "step": 9770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8335829036522005e-05,
+      "loss": 1.7329,
+      "step": 9771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.833549782605451e-05,
+      "loss": 1.7647,
+      "step": 9772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8335166585622873e-05,
+      "loss": 2.1411,
+      "step": 9773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.833483531522828e-05,
+      "loss": 1.662,
+      "step": 9774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8334504014871926e-05,
+      "loss": 1.6722,
+      "step": 9775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8334172684555004e-05,
+      "loss": 2.0011,
+      "step": 9776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.83338413242787e-05,
+      "loss": 2.1098,
+      "step": 9777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8333509934044208e-05,
+      "loss": 1.7294,
+      "step": 9778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.833317851385272e-05,
+      "loss": 1.8688,
+      "step": 9779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8332847063705423e-05,
+      "loss": 1.9625,
+      "step": 9780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.833251558360351e-05,
+      "loss": 2.0276,
+      "step": 9781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.833218407354818e-05,
+      "loss": 1.7506,
+      "step": 9782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8331852533540614e-05,
+      "loss": 1.635,
+      "step": 9783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8331520963582008e-05,
+      "loss": 2.2088,
+      "step": 9784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8331189363673557e-05,
+      "loss": 2.0866,
+      "step": 9785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8330857733816453e-05,
+      "loss": 1.989,
+      "step": 9786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.833052607401188e-05,
+      "loss": 1.7334,
+      "step": 9787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8330194384261036e-05,
+      "loss": 1.9174,
+      "step": 9788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8329862664565118e-05,
+      "loss": 2.157,
+      "step": 9789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8329530914925315e-05,
+      "loss": 2.1319,
+      "step": 9790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8329199135342812e-05,
+      "loss": 1.8389,
+      "step": 9791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.832886732581881e-05,
+      "loss": 1.6846,
+      "step": 9792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8328535486354502e-05,
+      "loss": 1.6317,
+      "step": 9793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8328203616951074e-05,
+      "loss": 1.7828,
+      "step": 9794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.832787171760973e-05,
+      "loss": 2.0719,
+      "step": 9795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8327539788331653e-05,
+      "loss": 1.7292,
+      "step": 9796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8327207829118037e-05,
+      "loss": 1.6794,
+      "step": 9797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.8326875839970082e-05,
+      "loss": 1.8507,
+      "step": 9798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8326543820888976e-05,
+      "loss": 1.9224,
+      "step": 9799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8326211771875915e-05,
+      "loss": 1.8055,
+      "step": 9800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.832587969293209e-05,
+      "loss": 2.1389,
+      "step": 9801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.83255475840587e-05,
+      "loss": 1.6091,
+      "step": 9802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8325215445256933e-05,
+      "loss": 1.6106,
+      "step": 9803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8324883276527985e-05,
+      "loss": 1.7949,
+      "step": 9804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8324551077873055e-05,
+      "loss": 1.6796,
+      "step": 9805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8324218849293328e-05,
+      "loss": 2.3488,
+      "step": 9806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8323886590790005e-05,
+      "loss": 2.1159,
+      "step": 9807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8323554302364273e-05,
+      "loss": 1.8699,
+      "step": 9808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8323221984017337e-05,
+      "loss": 1.7937,
+      "step": 9809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.8322889635750384e-05,
+      "loss": 2.5415,
+      "step": 9810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8322557257564612e-05,
+      "loss": 2.0038,
+      "step": 9811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8322224849461216e-05,
+      "loss": 1.6787,
+      "step": 9812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.832189241144139e-05,
+      "loss": 2.1409,
+      "step": 9813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8321559943506323e-05,
+      "loss": 1.9763,
+      "step": 9814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8321227445657224e-05,
+      "loss": 1.9157,
+      "step": 9815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.832089491789528e-05,
+      "loss": 1.9195,
+      "step": 9816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8320562360221677e-05,
+      "loss": 1.8358,
+      "step": 9817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.832022977263763e-05,
+      "loss": 1.649,
+      "step": 9818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.831989715514432e-05,
+      "loss": 2.0852,
+      "step": 9819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8319564507742948e-05,
+      "loss": 1.9164,
+      "step": 9820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.831923183043471e-05,
+      "loss": 1.6603,
+      "step": 9821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8318899123220802e-05,
+      "loss": 1.5988,
+      "step": 9822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8318566386102417e-05,
+      "loss": 1.7097,
+      "step": 9823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8318233619080756e-05,
+      "loss": 2.0472,
+      "step": 9824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.831790082215701e-05,
+      "loss": 2.1961,
+      "step": 9825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8317567995332376e-05,
+      "loss": 2.1039,
+      "step": 9826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8317235138608055e-05,
+      "loss": 2.1312,
+      "step": 9827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8316902251985237e-05,
+      "loss": 1.6697,
+      "step": 9828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8316569335465127e-05,
+      "loss": 2.0107,
+      "step": 9829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8316236389048915e-05,
+      "loss": 2.1009,
+      "step": 9830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.83159034127378e-05,
+      "loss": 2.0159,
+      "step": 9831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8315570406532978e-05,
+      "loss": 1.8701,
+      "step": 9832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8315237370435648e-05,
+      "loss": 1.7858,
+      "step": 9833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8314904304447007e-05,
+      "loss": 1.7975,
+      "step": 9834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.831457120856825e-05,
+      "loss": 1.7406,
+      "step": 9835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8314238082800573e-05,
+      "loss": 1.8655,
+      "step": 9836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.831390492714518e-05,
+      "loss": 1.7913,
+      "step": 9837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8313571741603266e-05,
+      "loss": 2.0095,
+      "step": 9838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8313238526176027e-05,
+      "loss": 1.6147,
+      "step": 9839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.831290528086466e-05,
+      "loss": 1.8712,
+      "step": 9840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8312572005670365e-05,
+      "loss": 2.0117,
+      "step": 9841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8312238700594342e-05,
+      "loss": 1.9448,
+      "step": 9842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.831190536563778e-05,
+      "loss": 1.9119,
+      "step": 9843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.831157200080189e-05,
+      "loss": 2.1473,
+      "step": 9844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.831123860608786e-05,
+      "loss": 1.9746,
+      "step": 9845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8310905181496895e-05,
+      "loss": 1.9037,
+      "step": 9846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8310571727030196e-05,
+      "loss": 1.7524,
+      "step": 9847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8310238242688953e-05,
+      "loss": 1.7861,
+      "step": 9848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8309904728474366e-05,
+      "loss": 1.7341,
+      "step": 9849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.830957118438764e-05,
+      "loss": 1.5196,
+      "step": 9850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8309237610429972e-05,
+      "loss": 1.8934,
+      "step": 9851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8308904006602555e-05,
+      "loss": 1.6777,
+      "step": 9852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8308570372906595e-05,
+      "loss": 1.9004,
+      "step": 9853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8308236709343294e-05,
+      "loss": 1.896,
+      "step": 9854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8307903015913838e-05,
+      "loss": 2.2602,
+      "step": 9855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8307569292619444e-05,
+      "loss": 1.9661,
+      "step": 9856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.83072355394613e-05,
+      "loss": 1.9768,
+      "step": 9857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8306901756440607e-05,
+      "loss": 1.8004,
+      "step": 9858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8306567943558567e-05,
+      "loss": 1.7609,
+      "step": 9859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8306234100816383e-05,
+      "loss": 1.8602,
+      "step": 9860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.830590022821525e-05,
+      "loss": 2.0576,
+      "step": 9861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.830556632575637e-05,
+      "loss": 1.8554,
+      "step": 9862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8305232393440942e-05,
+      "loss": 1.9822,
+      "step": 9863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.830489843127017e-05,
+      "loss": 1.7725,
+      "step": 9864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.8304564439245248e-05,
+      "loss": 2.1172,
+      "step": 9865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8304230417367386e-05,
+      "loss": 2.2402,
+      "step": 9866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8303896365637776e-05,
+      "loss": 1.7181,
+      "step": 9867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8303562284057627e-05,
+      "loss": 1.7996,
+      "step": 9868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.8303228172628128e-05,
+      "loss": 2.2198,
+      "step": 9869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8302894031350493e-05,
+      "loss": 2.032,
+      "step": 9870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8302559860225913e-05,
+      "loss": 1.8211,
+      "step": 9871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.83022256592556e-05,
+      "loss": 2.1666,
+      "step": 9872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8301891428440744e-05,
+      "loss": 2.0945,
+      "step": 9873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8301557167782556e-05,
+      "loss": 1.9109,
+      "step": 9874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.830122287728223e-05,
+      "loss": 1.1848,
+      "step": 9875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.830088855694097e-05,
+      "loss": 1.6233,
+      "step": 9876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.830055420675998e-05,
+      "loss": 1.7405,
+      "step": 9877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.830021982674046e-05,
+      "loss": 2.2319,
+      "step": 9878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8299885416883616e-05,
+      "loss": 1.629,
+      "step": 9879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8299550977190646e-05,
+      "loss": 1.7473,
+      "step": 9880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.829921650766275e-05,
+      "loss": 1.8546,
+      "step": 9881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8298882008301136e-05,
+      "loss": 2.0423,
+      "step": 9882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8298547479107e-05,
+      "loss": 2.2531,
+      "step": 9883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.8298212920081553e-05,
+      "loss": 1.8446,
+      "step": 9884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8297878331225993e-05,
+      "loss": 2.0671,
+      "step": 9885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.829754371254152e-05,
+      "loss": 1.904,
+      "step": 9886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.829720906402934e-05,
+      "loss": 2.2315,
+      "step": 9887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8296874385690655e-05,
+      "loss": 2.1084,
+      "step": 9888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.829653967752667e-05,
+      "loss": 1.7613,
+      "step": 9889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8296204939538588e-05,
+      "loss": 1.9547,
+      "step": 9890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.829587017172761e-05,
+      "loss": 1.8473,
+      "step": 9891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.829553537409494e-05,
+      "loss": 1.9464,
+      "step": 9892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.8295200546641783e-05,
+      "loss": 2.2207,
+      "step": 9893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8294865689369342e-05,
+      "loss": 1.7842,
+      "step": 9894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8294530802278817e-05,
+      "loss": 2.3219,
+      "step": 9895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8294195885371422e-05,
+      "loss": 1.8899,
+      "step": 9896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.829386093864835e-05,
+      "loss": 1.7795,
+      "step": 9897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.829352596211081e-05,
+      "loss": 1.737,
+      "step": 9898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8293190955760006e-05,
+      "loss": 1.7806,
+      "step": 9899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8292855919597144e-05,
+      "loss": 2.0995,
+      "step": 9900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8292520853623422e-05,
+      "loss": 2.3156,
+      "step": 9901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8292185757840052e-05,
+      "loss": 1.841,
+      "step": 9902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.829185063224823e-05,
+      "loss": 1.5651,
+      "step": 9903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.829151547684917e-05,
+      "loss": 1.8721,
+      "step": 9904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8291180291644073e-05,
+      "loss": 2.3931,
+      "step": 9905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8290845076634145e-05,
+      "loss": 1.8508,
+      "step": 9906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8290509831820586e-05,
+      "loss": 1.7575,
+      "step": 9907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8290174557204607e-05,
+      "loss": 2.0634,
+      "step": 9908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.828983925278741e-05,
+      "loss": 1.754,
+      "step": 9909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8289503918570202e-05,
+      "loss": 2.3429,
+      "step": 9910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.828916855455419e-05,
+      "loss": 1.8813,
+      "step": 9911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.8288833160740575e-05,
+      "loss": 1.8194,
+      "step": 9912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8288497737130563e-05,
+      "loss": 1.8009,
+      "step": 9913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8288162283725366e-05,
+      "loss": 1.8316,
+      "step": 9914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8287826800526182e-05,
+      "loss": 1.9697,
+      "step": 9915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8287491287534224e-05,
+      "loss": 2.06,
+      "step": 9916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8287155744750694e-05,
+      "loss": 1.5952,
+      "step": 9917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8286820172176794e-05,
+      "loss": 2.0421,
+      "step": 9918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.828648456981374e-05,
+      "loss": 2.0503,
+      "step": 9919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8286148937662734e-05,
+      "loss": 1.9289,
+      "step": 9920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8285813275724982e-05,
+      "loss": 2.1509,
+      "step": 9921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.828547758400169e-05,
+      "loss": 2.6116,
+      "step": 9922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8285141862494064e-05,
+      "loss": 2.1914,
+      "step": 9923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8284806111203318e-05,
+      "loss": 2.0057,
+      "step": 9924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8284470330130646e-05,
+      "loss": 1.8007,
+      "step": 9925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8284134519277268e-05,
+      "loss": 1.7977,
+      "step": 9926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8283798678644384e-05,
+      "loss": 1.8371,
+      "step": 9927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8283462808233202e-05,
+      "loss": 2.2375,
+      "step": 9928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.828312690804493e-05,
+      "loss": 2.0334,
+      "step": 9929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8282790978080775e-05,
+      "loss": 2.1837,
+      "step": 9930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8282455018341946e-05,
+      "loss": 2.3103,
+      "step": 9931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.828211902882965e-05,
+      "loss": 1.874,
+      "step": 9932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8281783009545094e-05,
+      "loss": 1.8,
+      "step": 9933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8281446960489487e-05,
+      "loss": 2.2115,
+      "step": 9934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.828111088166404e-05,
+      "loss": 1.9607,
+      "step": 9935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8280774773069954e-05,
+      "loss": 1.9957,
+      "step": 9936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.828043863470844e-05,
+      "loss": 1.812,
+      "step": 9937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.828010246658071e-05,
+      "loss": 1.8995,
+      "step": 9938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.827976626868797e-05,
+      "loss": 1.7369,
+      "step": 9939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8279430041031426e-05,
+      "loss": 1.8711,
+      "step": 9940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.827909378361229e-05,
+      "loss": 1.8061,
+      "step": 9941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.827875749643177e-05,
+      "loss": 1.4829,
+      "step": 9942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8278421179491078e-05,
+      "loss": 1.7874,
+      "step": 9943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8278084832791416e-05,
+      "loss": 1.9336,
+      "step": 9944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8277748456333997e-05,
+      "loss": 1.9486,
+      "step": 9945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8277412050120032e-05,
+      "loss": 1.9011,
+      "step": 9946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8277075614150725e-05,
+      "loss": 1.8766,
+      "step": 9947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.827673914842729e-05,
+      "loss": 1.9893,
+      "step": 9948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.827640265295094e-05,
+      "loss": 2.006,
+      "step": 9949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.827606612772287e-05,
+      "loss": 1.8347,
+      "step": 9950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8275729572744307e-05,
+      "loss": 1.7565,
+      "step": 9951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.827539298801645e-05,
+      "loss": 1.6745,
+      "step": 9952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8275056373540514e-05,
+      "loss": 1.8297,
+      "step": 9953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8274719729317707e-05,
+      "loss": 1.9855,
+      "step": 9954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.827438305534924e-05,
+      "loss": 1.9356,
+      "step": 9955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.827404635163632e-05,
+      "loss": 2.0362,
+      "step": 9956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.827370961818016e-05,
+      "loss": 1.7171,
+      "step": 9957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8273372854981976e-05,
+      "loss": 1.9225,
+      "step": 9958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.827303606204297e-05,
+      "loss": 2.1428,
+      "step": 9959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8272699239364354e-05,
+      "loss": 1.9747,
+      "step": 9960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8272362386947345e-05,
+      "loss": 1.7771,
+      "step": 9961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8272025504793146e-05,
+      "loss": 2.1976,
+      "step": 9962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8271688592902972e-05,
+      "loss": 1.8636,
+      "step": 9963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8271351651278034e-05,
+      "loss": 2.2939,
+      "step": 9964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8271014679919544e-05,
+      "loss": 1.9057,
+      "step": 9965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8270677678828712e-05,
+      "loss": 1.7674,
+      "step": 9966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8270340648006747e-05,
+      "loss": 1.7582,
+      "step": 9967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8270003587454867e-05,
+      "loss": 2.0083,
+      "step": 9968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8269666497174275e-05,
+      "loss": 1.9435,
+      "step": 9969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8269329377166195e-05,
+      "loss": 1.9404,
+      "step": 9970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8268992227431828e-05,
+      "loss": 1.7525,
+      "step": 9971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8268655047972386e-05,
+      "loss": 1.787,
+      "step": 9972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.826831783878909e-05,
+      "loss": 2.2797,
+      "step": 9973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.826798059988314e-05,
+      "loss": 1.6779,
+      "step": 9974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8267643331255763e-05,
+      "loss": 2.0998,
+      "step": 9975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8267306032908162e-05,
+      "loss": 1.9213,
+      "step": 9976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.826696870484155e-05,
+      "loss": 1.7107,
+      "step": 9977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8266631347057138e-05,
+      "loss": 2.0369,
+      "step": 9978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.826629395955614e-05,
+      "loss": 1.9842,
+      "step": 9979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.8265956542339774e-05,
+      "loss": 4.1365,
+      "step": 9980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8265619095409248e-05,
+      "loss": 1.871,
+      "step": 9981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8265281618765775e-05,
+      "loss": 1.6977,
+      "step": 9982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.826494411241057e-05,
+      "loss": 1.6157,
+      "step": 9983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8264606576344847e-05,
+      "loss": 1.6523,
+      "step": 9984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.826426901056982e-05,
+      "loss": 1.6485,
+      "step": 9985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8263931415086695e-05,
+      "loss": 1.7903,
+      "step": 9986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8263593789896692e-05,
+      "loss": 2.1607,
+      "step": 9987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.8263256135001024e-05,
+      "loss": 1.8363,
+      "step": 9988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8262918450400903e-05,
+      "loss": 1.9422,
+      "step": 9989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.826258073609755e-05,
+      "loss": 1.9381,
+      "step": 9990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.826224299209217e-05,
+      "loss": 1.7593,
+      "step": 9991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8261905218385977e-05,
+      "loss": 1.6594,
+      "step": 9992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8261567414980192e-05,
+      "loss": 1.6912,
+      "step": 9993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8261229581876025e-05,
+      "loss": 2.0313,
+      "step": 9994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8260891719074694e-05,
+      "loss": 1.817,
+      "step": 9995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.826055382657741e-05,
+      "loss": 1.7859,
+      "step": 9996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8260215904385385e-05,
+      "loss": 2.0457,
+      "step": 9997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.825987795249984e-05,
+      "loss": 1.716,
+      "step": 9998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.825953997092199e-05,
+      "loss": 1.9469,
+      "step": 9999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.825920195965304e-05,
+      "loss": 1.9902,
+      "step": 10000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.825886391869422e-05,
+      "loss": 2.1375,
+      "step": 10001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.825852584804673e-05,
+      "loss": 1.8434,
+      "step": 10002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8258187747711797e-05,
+      "loss": 2.4593,
+      "step": 10003
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.8257849617690632e-05,
+      "loss": 1.8147,
+      "step": 10004
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.825751145798445e-05,
+      "loss": 1.57,
+      "step": 10005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8257173268594468e-05,
+      "loss": 1.8893,
+      "step": 10006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.82568350495219e-05,
+      "loss": 1.929,
+      "step": 10007
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8256496800767963e-05,
+      "loss": 1.8797,
+      "step": 10008
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.825615852233387e-05,
+      "loss": 1.558,
+      "step": 10009
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8255820214220843e-05,
+      "loss": 2.103,
+      "step": 10010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8255481876430093e-05,
+      "loss": 1.8624,
+      "step": 10011
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.825514350896284e-05,
+      "loss": 1.6237,
+      "step": 10012
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8254805111820296e-05,
+      "loss": 1.7809,
+      "step": 10013
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8254466685003683e-05,
+      "loss": 1.8173,
+      "step": 10014
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.825412822851421e-05,
+      "loss": 1.85,
+      "step": 10015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.82537897423531e-05,
+      "loss": 2.0295,
+      "step": 10016
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.825345122652157e-05,
+      "loss": 1.8379,
+      "step": 10017
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.825311268102083e-05,
+      "loss": 1.7328,
+      "step": 10018
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8252774105852106e-05,
+      "loss": 1.976,
+      "step": 10019
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.825243550101661e-05,
+      "loss": 1.736,
+      "step": 10020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.825209686651556e-05,
+      "loss": 1.9508,
+      "step": 10021
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8251758202350173e-05,
+      "loss": 1.6006,
+      "step": 10022
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8251419508521666e-05,
+      "loss": 1.9963,
+      "step": 10023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8251080785031258e-05,
+      "loss": 2.0038,
+      "step": 10024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8250742031880168e-05,
+      "loss": 1.8312,
+      "step": 10025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8250403249069607e-05,
+      "loss": 2.077,
+      "step": 10026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.82500644366008e-05,
+      "loss": 1.9743,
+      "step": 10027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8249725594474964e-05,
+      "loss": 1.6244,
+      "step": 10028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8249386722693312e-05,
+      "loss": 2.0973,
+      "step": 10029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8249047821257067e-05,
+      "loss": 1.689,
+      "step": 10030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8248708890167445e-05,
+      "loss": 1.6431,
+      "step": 10031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.824836992942567e-05,
+      "loss": 1.7517,
+      "step": 10032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8248030939032954e-05,
+      "loss": 1.78,
+      "step": 10033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8247691918990514e-05,
+      "loss": 1.9166,
+      "step": 10034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8247352869299573e-05,
+      "loss": 1.7017,
+      "step": 10035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.824701378996135e-05,
+      "loss": 2.1371,
+      "step": 10036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8246674680977062e-05,
+      "loss": 2.1576,
+      "step": 10037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.824633554234793e-05,
+      "loss": 2.0654,
+      "step": 10038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.824599637407517e-05,
+      "loss": 1.8487,
+      "step": 10039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8245657176160006e-05,
+      "loss": 1.8954,
+      "step": 10040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8245317948603652e-05,
+      "loss": 1.7374,
+      "step": 10041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8244978691407332e-05,
+      "loss": 1.7237,
+      "step": 10042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.824463940457226e-05,
+      "loss": 1.8704,
+      "step": 10043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8244300088099664e-05,
+      "loss": 1.8269,
+      "step": 10044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8243960741990756e-05,
+      "loss": 2.3881,
+      "step": 10045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8243621366246757e-05,
+      "loss": 1.942,
+      "step": 10046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.824328196086889e-05,
+      "loss": 1.8707,
+      "step": 10047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.824294252585838e-05,
+      "loss": 1.9486,
+      "step": 10048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8242603061216432e-05,
+      "loss": 1.72,
+      "step": 10049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.824226356694428e-05,
+      "loss": 1.7797,
+      "step": 10050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.824192404304314e-05,
+      "loss": 1.7625,
+      "step": 10051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.824158448951423e-05,
+      "loss": 1.7096,
+      "step": 10052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.8241244906358774e-05,
+      "loss": 1.4157,
+      "step": 10053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.824090529357799e-05,
+      "loss": 2.1989,
+      "step": 10054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8240565651173103e-05,
+      "loss": 1.9968,
+      "step": 10055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.824022597914533e-05,
+      "loss": 2.1557,
+      "step": 10056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8239886277495894e-05,
+      "loss": 1.7221,
+      "step": 10057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8239546546226014e-05,
+      "loss": 2.0649,
+      "step": 10058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8239206785336913e-05,
+      "loss": 1.5299,
+      "step": 10059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8238866994829815e-05,
+      "loss": 1.6227,
+      "step": 10060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8238527174705938e-05,
+      "loss": 1.9109,
+      "step": 10061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.82381873249665e-05,
+      "loss": 1.9417,
+      "step": 10062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.823784744561273e-05,
+      "loss": 1.5978,
+      "step": 10063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8237507536645847e-05,
+      "loss": 1.9667,
+      "step": 10064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8237167598067072e-05,
+      "loss": 1.9902,
+      "step": 10065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.823682762987763e-05,
+      "loss": 1.8266,
+      "step": 10066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8236487632078735e-05,
+      "loss": 1.8463,
+      "step": 10067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.823614760467162e-05,
+      "loss": 1.9938,
+      "step": 10068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.82358075476575e-05,
+      "loss": 2.1136,
+      "step": 10069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.82354674610376e-05,
+      "loss": 1.6856,
+      "step": 10070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8235127344813138e-05,
+      "loss": 1.7786,
+      "step": 10071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8234787198985344e-05,
+      "loss": 1.9051,
+      "step": 10072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.823444702355544e-05,
+      "loss": 1.7944,
+      "step": 10073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8234106818524642e-05,
+      "loss": 1.8201,
+      "step": 10074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8233766583894178e-05,
+      "loss": 1.9691,
+      "step": 10075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8233426319665272e-05,
+      "loss": 2.3136,
+      "step": 10076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8233086025839145e-05,
+      "loss": 1.8715,
+      "step": 10077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8232745702417022e-05,
+      "loss": 2.2217,
+      "step": 10078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8232405349400124e-05,
+      "loss": 1.7222,
+      "step": 10079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8232064966789677e-05,
+      "loss": 1.9831,
+      "step": 10080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.82317245545869e-05,
+      "loss": 1.8508,
+      "step": 10081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8231384112793024e-05,
+      "loss": 1.9276,
+      "step": 10082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.823104364140927e-05,
+      "loss": 2.0794,
+      "step": 10083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8230703140436853e-05,
+      "loss": 1.7354,
+      "step": 10084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.823036260987701e-05,
+      "loss": 2.1475,
+      "step": 10085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.823002204973096e-05,
+      "loss": 1.7673,
+      "step": 10086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8229681459999925e-05,
+      "loss": 1.7446,
+      "step": 10087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.822934084068513e-05,
+      "loss": 2.2267,
+      "step": 10088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8229000191787807e-05,
+      "loss": 1.8684,
+      "step": 10089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.822865951330917e-05,
+      "loss": 1.9406,
+      "step": 10090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8228318805250447e-05,
+      "loss": 2.2197,
+      "step": 10091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.822797806761287e-05,
+      "loss": 1.7786,
+      "step": 10092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8227637300397652e-05,
+      "loss": 1.802,
+      "step": 10093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8227296503606027e-05,
+      "loss": 2.5022,
+      "step": 10094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8226955677239213e-05,
+      "loss": 2.052,
+      "step": 10095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8226614821298444e-05,
+      "loss": 1.845,
+      "step": 10096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8226273935784936e-05,
+      "loss": 2.1461,
+      "step": 10097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.822593302069992e-05,
+      "loss": 1.8204,
+      "step": 10098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8225592076044618e-05,
+      "loss": 1.8031,
+      "step": 10099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.822525110182026e-05,
+      "loss": 2.0456,
+      "step": 10100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.822491009802807e-05,
+      "loss": 1.7648,
+      "step": 10101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.822456906466927e-05,
+      "loss": 2.0067,
+      "step": 10102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8224228001745095e-05,
+      "loss": 1.6166,
+      "step": 10103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8223886909256762e-05,
+      "loss": 1.8876,
+      "step": 10104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.82235457872055e-05,
+      "loss": 1.5956,
+      "step": 10105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8223204635592537e-05,
+      "loss": 1.8237,
+      "step": 10106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.82228634544191e-05,
+      "loss": 1.8938,
+      "step": 10107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.822252224368641e-05,
+      "loss": 1.8634,
+      "step": 10108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.82221810033957e-05,
+      "loss": 1.894,
+      "step": 10109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.822183973354819e-05,
+      "loss": 2.028,
+      "step": 10110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8221498434145114e-05,
+      "loss": 1.902,
+      "step": 10111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8221157105187692e-05,
+      "loss": 1.8333,
+      "step": 10112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8220815746677156e-05,
+      "loss": 1.9042,
+      "step": 10113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8220474358614734e-05,
+      "loss": 1.9583,
+      "step": 10114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8220132941001647e-05,
+      "loss": 1.6799,
+      "step": 10115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8219791493839128e-05,
+      "loss": 1.9246,
+      "step": 10116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8219450017128403e-05,
+      "loss": 1.6748,
+      "step": 10117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.82191085108707e-05,
+      "loss": 2.0151,
+      "step": 10118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.8218766975067242e-05,
+      "loss": 2.3913,
+      "step": 10119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8218425409719262e-05,
+      "loss": 1.9072,
+      "step": 10120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8218083814827987e-05,
+      "loss": 1.7238,
+      "step": 10121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8217742190394645e-05,
+      "loss": 1.806,
+      "step": 10122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8217400536420463e-05,
+      "loss": 1.7161,
+      "step": 10123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8217058852906667e-05,
+      "loss": 2.1396,
+      "step": 10124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.821671713985449e-05,
+      "loss": 2.0077,
+      "step": 10125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8216375397265156e-05,
+      "loss": 1.7269,
+      "step": 10126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8216033625139897e-05,
+      "loss": 1.4777,
+      "step": 10127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.821569182347994e-05,
+      "loss": 2.054,
+      "step": 10128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8215349992286517e-05,
+      "loss": 1.7224,
+      "step": 10129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.821500813156085e-05,
+      "loss": 1.7237,
+      "step": 10130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.821466624130417e-05,
+      "loss": 1.8561,
+      "step": 10131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.821432432151771e-05,
+      "loss": 2.0641,
+      "step": 10132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8213982372202695e-05,
+      "loss": 1.747,
+      "step": 10133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.821364039336036e-05,
+      "loss": 1.899,
+      "step": 10134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8213298384991926e-05,
+      "loss": 1.9496,
+      "step": 10135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8212956347098628e-05,
+      "loss": 1.9748,
+      "step": 10136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8212614279681698e-05,
+      "loss": 1.9856,
+      "step": 10137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8212272182742356e-05,
+      "loss": 1.7518,
+      "step": 10138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8211930056281842e-05,
+      "loss": 1.7252,
+      "step": 10139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.821158790030138e-05,
+      "loss": 1.6459,
+      "step": 10140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.82112457148022e-05,
+      "loss": 1.7677,
+      "step": 10141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8210903499785533e-05,
+      "loss": 2.3393,
+      "step": 10142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.821056125525261e-05,
+      "loss": 2.0592,
+      "step": 10143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8210218981204667e-05,
+      "loss": 1.7657,
+      "step": 10144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.820987667764292e-05,
+      "loss": 2.1423,
+      "step": 10145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8209534344568614e-05,
+      "loss": 1.5028,
+      "step": 10146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.820919198198297e-05,
+      "loss": 1.8499,
+      "step": 10147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8208849589887224e-05,
+      "loss": 1.9365,
+      "step": 10148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8208507168282607e-05,
+      "loss": 2.0516,
+      "step": 10149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8208164717170343e-05,
+      "loss": 1.7277,
+      "step": 10150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.820782223655167e-05,
+      "loss": 2.0141,
+      "step": 10151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.820747972642782e-05,
+      "loss": 1.6372,
+      "step": 10152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.820713718680002e-05,
+      "loss": 1.9808,
+      "step": 10153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.82067946176695e-05,
+      "loss": 1.8311,
+      "step": 10154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8206452019037497e-05,
+      "loss": 2.1584,
+      "step": 10155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.820610939090524e-05,
+      "loss": 1.6671,
+      "step": 10156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8205766733273957e-05,
+      "loss": 2.1161,
+      "step": 10157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.8205424046144887e-05,
+      "loss": 1.7125,
+      "step": 10158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8205081329519257e-05,
+      "loss": 1.7308,
+      "step": 10159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.82047385833983e-05,
+      "loss": 1.8306,
+      "step": 10160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8204395807783247e-05,
+      "loss": 1.7455,
+      "step": 10161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8204053002675332e-05,
+      "loss": 2.0217,
+      "step": 10162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8203710168075786e-05,
+      "loss": 1.945,
+      "step": 10163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8203367303985843e-05,
+      "loss": 2.0037,
+      "step": 10164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8203024410406732e-05,
+      "loss": 1.7288,
+      "step": 10165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8202681487339694e-05,
+      "loss": 1.7435,
+      "step": 10166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.820233853478595e-05,
+      "loss": 2.1155,
+      "step": 10167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8201995552746745e-05,
+      "loss": 1.9771,
+      "step": 10168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8201652541223304e-05,
+      "loss": 1.6255,
+      "step": 10169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8201309500216858e-05,
+      "loss": 2.266,
+      "step": 10170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8200966429728645e-05,
+      "loss": 1.736,
+      "step": 10171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8200623329759897e-05,
+      "loss": 2.0892,
+      "step": 10172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.820028020031185e-05,
+      "loss": 1.6903,
+      "step": 10173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8199937041385735e-05,
+      "loss": 2.0728,
+      "step": 10174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8199593852982783e-05,
+      "loss": 1.6698,
+      "step": 10175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.819925063510423e-05,
+      "loss": 1.915,
+      "step": 10176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.819890738775131e-05,
+      "loss": 1.8379,
+      "step": 10177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.819856411092526e-05,
+      "loss": 1.9823,
+      "step": 10178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.819822080462731e-05,
+      "loss": 2.1891,
+      "step": 10179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.819787746885869e-05,
+      "loss": 1.8877,
+      "step": 10180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8197534103620647e-05,
+      "loss": 2.2338,
+      "step": 10181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.81971907089144e-05,
+      "loss": 2.0147,
+      "step": 10182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8196847284741198e-05,
+      "loss": 1.7946,
+      "step": 10183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8196503831102266e-05,
+      "loss": 2.0454,
+      "step": 10184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.375,
+      "learning_rate": 1.819616034799884e-05,
+      "loss": 2.0053,
+      "step": 10185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8195816835432157e-05,
+      "loss": 1.6576,
+      "step": 10186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.819547329340345e-05,
+      "loss": 1.7071,
+      "step": 10187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8195129721913957e-05,
+      "loss": 1.6999,
+      "step": 10188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8194786120964906e-05,
+      "loss": 1.8564,
+      "step": 10189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.819444249055754e-05,
+      "loss": 1.7675,
+      "step": 10190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8194098830693093e-05,
+      "loss": 1.8413,
+      "step": 10191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8193755141372796e-05,
+      "loss": 2.0597,
+      "step": 10192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8193411422597888e-05,
+      "loss": 2.1526,
+      "step": 10193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.81930676743696e-05,
+      "loss": 1.9197,
+      "step": 10194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8192723896689175e-05,
+      "loss": 1.9691,
+      "step": 10195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.8192380089557845e-05,
+      "loss": 1.8286,
+      "step": 10196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8192036252976842e-05,
+      "loss": 1.8573,
+      "step": 10197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.819169238694741e-05,
+      "loss": 1.193,
+      "step": 10198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.819134849147078e-05,
+      "loss": 1.9737,
+      "step": 10199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8191004566548192e-05,
+      "loss": 2.0196,
+      "step": 10200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8190660612180878e-05,
+      "loss": 1.564,
+      "step": 10201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8190316628370075e-05,
+      "loss": 1.8403,
+      "step": 10202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.818997261511702e-05,
+      "loss": 1.6748,
+      "step": 10203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.8189628572422954e-05,
+      "loss": 1.9236,
+      "step": 10204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.818928450028911e-05,
+      "loss": 1.9294,
+      "step": 10205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8188940398716722e-05,
+      "loss": 1.8376,
+      "step": 10206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.818859626770703e-05,
+      "loss": 1.8556,
+      "step": 10207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.818825210726127e-05,
+      "loss": 1.8141,
+      "step": 10208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8187907917380682e-05,
+      "loss": 1.8543,
+      "step": 10209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8187563698066503e-05,
+      "loss": 2.0043,
+      "step": 10210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.818721944931997e-05,
+      "loss": 2.0381,
+      "step": 10211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8186875171142314e-05,
+      "loss": 1.9922,
+      "step": 10212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.818653086353478e-05,
+      "loss": 2.0373,
+      "step": 10213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8186186526498604e-05,
+      "loss": 1.9565,
+      "step": 10214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8185842160035023e-05,
+      "loss": 1.8272,
+      "step": 10215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8185497764145273e-05,
+      "loss": 1.8607,
+      "step": 10216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.81851533388306e-05,
+      "loss": 1.8538,
+      "step": 10217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8184808884092232e-05,
+      "loss": 2.1725,
+      "step": 10218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8184464399931414e-05,
+      "loss": 1.7608,
+      "step": 10219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.818411988634938e-05,
+      "loss": 2.1076,
+      "step": 10220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8183775343347372e-05,
+      "loss": 1.9215,
+      "step": 10221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8183430770926625e-05,
+      "loss": 1.8158,
+      "step": 10222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8183086169088383e-05,
+      "loss": 2.0502,
+      "step": 10223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.818274153783388e-05,
+      "loss": 2.0069,
+      "step": 10224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8182396877164355e-05,
+      "loss": 1.9186,
+      "step": 10225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.818205218708105e-05,
+      "loss": 1.6494,
+      "step": 10226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8181707467585204e-05,
+      "loss": 1.6348,
+      "step": 10227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.818136271867805e-05,
+      "loss": 1.827,
+      "step": 10228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8181017940360837e-05,
+      "loss": 1.8731,
+      "step": 10229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8180673132634795e-05,
+      "loss": 2.1272,
+      "step": 10230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8180328295501168e-05,
+      "loss": 1.6929,
+      "step": 10231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.81799834289612e-05,
+      "loss": 1.937,
+      "step": 10232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.817963853301612e-05,
+      "loss": 1.9125,
+      "step": 10233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8179293607667177e-05,
+      "loss": 1.8751,
+      "step": 10234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.817894865291561e-05,
+      "loss": 1.8762,
+      "step": 10235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8178603668762653e-05,
+      "loss": 1.7674,
+      "step": 10236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.817825865520955e-05,
+      "loss": 1.7814,
+      "step": 10237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8177913612257545e-05,
+      "loss": 1.9564,
+      "step": 10238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8177568539907873e-05,
+      "loss": 1.6818,
+      "step": 10239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8177223438161774e-05,
+      "loss": 1.9025,
+      "step": 10240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.817687830702049e-05,
+      "loss": 1.9516,
+      "step": 10241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8176533146485266e-05,
+      "loss": 1.6212,
+      "step": 10242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.8176187956557338e-05,
+      "loss": 1.7707,
+      "step": 10243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8175842737237946e-05,
+      "loss": 1.7323,
+      "step": 10244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8175497488528333e-05,
+      "loss": 1.7816,
+      "step": 10245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.817515221042974e-05,
+      "loss": 1.8373,
+      "step": 10246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.817480690294341e-05,
+      "loss": 1.8554,
+      "step": 10247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8174461566070584e-05,
+      "loss": 1.8591,
+      "step": 10248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.81741161998125e-05,
+      "loss": 1.7186,
+      "step": 10249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.81737708041704e-05,
+      "loss": 2.1368,
+      "step": 10250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8173425379145526e-05,
+      "loss": 1.6524,
+      "step": 10251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8173079924739126e-05,
+      "loss": 2.0576,
+      "step": 10252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.817273444095243e-05,
+      "loss": 1.9303,
+      "step": 10253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.817238892778669e-05,
+      "loss": 1.7587,
+      "step": 10254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8172043385243147e-05,
+      "loss": 1.9608,
+      "step": 10255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8171697813323038e-05,
+      "loss": 1.8993,
+      "step": 10256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8171352212027606e-05,
+      "loss": 1.8803,
+      "step": 10257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8171006581358098e-05,
+      "loss": 1.4945,
+      "step": 10258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.817066092131575e-05,
+      "loss": 1.818,
+      "step": 10259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.817031523190181e-05,
+      "loss": 2.1782,
+      "step": 10260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8169969513117522e-05,
+      "loss": 2.1962,
+      "step": 10261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.8169623764964124e-05,
+      "loss": 1.8473,
+      "step": 10262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.816927798744286e-05,
+      "loss": 1.9374,
+      "step": 10263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8168932180554975e-05,
+      "loss": 1.7709,
+      "step": 10264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.816858634430171e-05,
+      "loss": 1.8445,
+      "step": 10265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8168240478684308e-05,
+      "loss": 1.8894,
+      "step": 10266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8167894583704014e-05,
+      "loss": 1.793,
+      "step": 10267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8167548659362067e-05,
+      "loss": 1.65,
+      "step": 10268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.816720270565972e-05,
+      "loss": 2.2725,
+      "step": 10269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.816685672259821e-05,
+      "loss": 1.7829,
+      "step": 10270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8166510710178776e-05,
+      "loss": 1.8015,
+      "step": 10271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8166164668402674e-05,
+      "loss": 1.9256,
+      "step": 10272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8165818597271138e-05,
+      "loss": 2.1807,
+      "step": 10273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8165472496785414e-05,
+      "loss": 1.6394,
+      "step": 10274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.8165126366946753e-05,
+      "loss": 1.9041,
+      "step": 10275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8164780207756387e-05,
+      "loss": 1.6629,
+      "step": 10276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.816443401921557e-05,
+      "loss": 1.8939,
+      "step": 10277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8164087801325544e-05,
+      "loss": 1.7739,
+      "step": 10278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8163741554087554e-05,
+      "loss": 1.8012,
+      "step": 10279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8163395277502843e-05,
+      "loss": 2.1994,
+      "step": 10280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8163048971572656e-05,
+      "loss": 1.8483,
+      "step": 10281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8162702636298238e-05,
+      "loss": 1.8411,
+      "step": 10282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.816235627168083e-05,
+      "loss": 1.8265,
+      "step": 10283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.816200987772169e-05,
+      "loss": 1.8793,
+      "step": 10284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.816166345442205e-05,
+      "loss": 1.9148,
+      "step": 10285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8161317001783162e-05,
+      "loss": 1.5335,
+      "step": 10286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.816097051980627e-05,
+      "loss": 1.7028,
+      "step": 10287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8160624008492616e-05,
+      "loss": 2.0127,
+      "step": 10288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8160277467843453e-05,
+      "loss": 1.9752,
+      "step": 10289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8159930897860017e-05,
+      "loss": 1.8505,
+      "step": 10290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8159584298543563e-05,
+      "loss": 1.7988,
+      "step": 10291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.815923766989533e-05,
+      "loss": 1.5397,
+      "step": 10292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.375,
+      "learning_rate": 1.815889101191657e-05,
+      "loss": 1.6296,
+      "step": 10293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8158544324608524e-05,
+      "loss": 1.8748,
+      "step": 10294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.8158197607972443e-05,
+      "loss": 1.8498,
+      "step": 10295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.815785086200957e-05,
+      "loss": 1.9226,
+      "step": 10296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8157504086721155e-05,
+      "loss": 2.2233,
+      "step": 10297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8157157282108438e-05,
+      "loss": 1.848,
+      "step": 10298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.815681044817267e-05,
+      "loss": 2.0565,
+      "step": 10299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8156463584915102e-05,
+      "loss": 1.7887,
+      "step": 10300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.815611669233697e-05,
+      "loss": 1.7903,
+      "step": 10301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.815576977043953e-05,
+      "loss": 2.0706,
+      "step": 10302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8155422819224028e-05,
+      "loss": 1.7886,
+      "step": 10303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8155075838691706e-05,
+      "loss": 1.5869,
+      "step": 10304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.815472882884382e-05,
+      "loss": 1.7876,
+      "step": 10305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.815438178968161e-05,
+      "loss": 1.8477,
+      "step": 10306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8154034721206327e-05,
+      "loss": 1.9955,
+      "step": 10307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8153687623419217e-05,
+      "loss": 1.4166,
+      "step": 10308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8153340496321524e-05,
+      "loss": 2.0586,
+      "step": 10309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8152993339914503e-05,
+      "loss": 1.7189,
+      "step": 10310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.81526461541994e-05,
+      "loss": 1.8863,
+      "step": 10311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8152298939177464e-05,
+      "loss": 1.9005,
+      "step": 10312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8151951694849937e-05,
+      "loss": 1.7258,
+      "step": 10313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8151604421218075e-05,
+      "loss": 2.2034,
+      "step": 10314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8151257118283122e-05,
+      "loss": 1.9055,
+      "step": 10315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.815090978604633e-05,
+      "loss": 2.6375,
+      "step": 10316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.815056242450894e-05,
+      "loss": 1.5863,
+      "step": 10317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8150215033672208e-05,
+      "loss": 1.9381,
+      "step": 10318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.814986761353738e-05,
+      "loss": 1.5249,
+      "step": 10319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8149520164105702e-05,
+      "loss": 1.6094,
+      "step": 10320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.814917268537843e-05,
+      "loss": 1.9422,
+      "step": 10321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.814882517735681e-05,
+      "loss": 2.2819,
+      "step": 10322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.814847764004209e-05,
+      "loss": 1.6681,
+      "step": 10323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8148130073435522e-05,
+      "loss": 1.7994,
+      "step": 10324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8147782477538348e-05,
+      "loss": 1.9229,
+      "step": 10325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8147434852351828e-05,
+      "loss": 2.088,
+      "step": 10326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.8147087197877202e-05,
+      "loss": 2.1376,
+      "step": 10327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8146739514115728e-05,
+      "loss": 1.8524,
+      "step": 10328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8146391801068652e-05,
+      "loss": 1.8819,
+      "step": 10329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.814604405873722e-05,
+      "loss": 1.7963,
+      "step": 10330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8145696287122692e-05,
+      "loss": 1.9158,
+      "step": 10331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.814534848622631e-05,
+      "loss": 1.6665,
+      "step": 10332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.814500065604932e-05,
+      "loss": 2.1053,
+      "step": 10333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8144652796592987e-05,
+      "loss": 2.1565,
+      "step": 10334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8144304907858552e-05,
+      "loss": 1.902,
+      "step": 10335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8143956989847267e-05,
+      "loss": 1.993,
+      "step": 10336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8143609042560378e-05,
+      "loss": 1.8492,
+      "step": 10337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8143261065999147e-05,
+      "loss": 2.0959,
+      "step": 10338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8142913060164816e-05,
+      "loss": 2.0454,
+      "step": 10339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8142565025058635e-05,
+      "loss": 2.0874,
+      "step": 10340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8142216960681863e-05,
+      "loss": 1.9781,
+      "step": 10341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8141868867035745e-05,
+      "loss": 1.7931,
+      "step": 10342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.814152074412153e-05,
+      "loss": 1.7915,
+      "step": 10343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.814117259194048e-05,
+      "loss": 2.0901,
+      "step": 10344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8140824410493833e-05,
+      "loss": 2.0545,
+      "step": 10345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8140476199782852e-05,
+      "loss": 1.9466,
+      "step": 10346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8140127959808784e-05,
+      "loss": 2.0554,
+      "step": 10347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.813977969057288e-05,
+      "loss": 2.5898,
+      "step": 10348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.8139431392076395e-05,
+      "loss": 1.6328,
+      "step": 10349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8139083064320576e-05,
+      "loss": 2.3191,
+      "step": 10350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.813873470730668e-05,
+      "loss": 2.1469,
+      "step": 10351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8138386321035957e-05,
+      "loss": 1.986,
+      "step": 10352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.813803790550966e-05,
+      "loss": 2.1147,
+      "step": 10353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.813768946072904e-05,
+      "loss": 1.9302,
+      "step": 10354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8137340986695354e-05,
+      "loss": 1.8238,
+      "step": 10355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.813699248340985e-05,
+      "loss": 1.6795,
+      "step": 10356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8136643950873782e-05,
+      "loss": 1.7605,
+      "step": 10357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8136295389088404e-05,
+      "loss": 2.1782,
+      "step": 10358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8135946798054967e-05,
+      "loss": 1.5848,
+      "step": 10359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8135598177774727e-05,
+      "loss": 1.6736,
+      "step": 10360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8135249528248936e-05,
+      "loss": 1.9708,
+      "step": 10361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.813490084947884e-05,
+      "loss": 2.129,
+      "step": 10362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.813455214146571e-05,
+      "loss": 2.1423,
+      "step": 10363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.25,
+      "learning_rate": 1.813420340421078e-05,
+      "loss": 2.0038,
+      "step": 10364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8133854637715317e-05,
+      "loss": 1.7917,
+      "step": 10365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.813350584198057e-05,
+      "loss": 1.7676,
+      "step": 10366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.813315701700779e-05,
+      "loss": 1.8091,
+      "step": 10367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.813280816279824e-05,
+      "loss": 2.0118,
+      "step": 10368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8132459279353164e-05,
+      "loss": 1.9344,
+      "step": 10369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.813211036667382e-05,
+      "loss": 1.7258,
+      "step": 10370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.813176142476146e-05,
+      "loss": 2.1213,
+      "step": 10371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8131412453617345e-05,
+      "loss": 1.828,
+      "step": 10372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8131063453242724e-05,
+      "loss": 1.909,
+      "step": 10373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8130714423638854e-05,
+      "loss": 1.8857,
+      "step": 10374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8130365364806986e-05,
+      "loss": 1.9467,
+      "step": 10375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.813001627674838e-05,
+      "loss": 1.7727,
+      "step": 10376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8129667159464283e-05,
+      "loss": 1.8694,
+      "step": 10377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.812931801295596e-05,
+      "loss": 2.0674,
+      "step": 10378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8128968837224663e-05,
+      "loss": 1.8837,
+      "step": 10379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.812861963227164e-05,
+      "loss": 1.6259,
+      "step": 10380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8128270398098155e-05,
+      "loss": 2.0586,
+      "step": 10381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.812792113470546e-05,
+      "loss": 2.072,
+      "step": 10382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8127571842094813e-05,
+      "loss": 1.8639,
+      "step": 10383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8127222520267462e-05,
+      "loss": 2.1198,
+      "step": 10384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.812687316922467e-05,
+      "loss": 1.7188,
+      "step": 10385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8126523788967693e-05,
+      "loss": 2.19,
+      "step": 10386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.8126174379497782e-05,
+      "loss": 1.6983,
+      "step": 10387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.8125824940816195e-05,
+      "loss": 2.0082,
+      "step": 10388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.8125475472924193e-05,
+      "loss": 1.8056,
+      "step": 10389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8125125975823028e-05,
+      "loss": 1.6847,
+      "step": 10390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8124776449513955e-05,
+      "loss": 1.7859,
+      "step": 10391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8124426893998234e-05,
+      "loss": 2.3804,
+      "step": 10392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8124077309277118e-05,
+      "loss": 1.692,
+      "step": 10393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.8123727695351862e-05,
+      "loss": 1.9055,
+      "step": 10394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.812337805222373e-05,
+      "loss": 1.9134,
+      "step": 10395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.8123028379893978e-05,
+      "loss": 1.6999,
+      "step": 10396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8122678678363854e-05,
+      "loss": 1.8999,
+      "step": 10397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8122328947634624e-05,
+      "loss": 1.7363,
+      "step": 10398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8121979187707542e-05,
+      "loss": 1.6107,
+      "step": 10399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8121629398583868e-05,
+      "loss": 1.5812,
+      "step": 10400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.812127958026485e-05,
+      "loss": 1.8288,
+      "step": 10401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.812092973275176e-05,
+      "loss": 1.8,
+      "step": 10402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8120579856045845e-05,
+      "loss": 1.8715,
+      "step": 10403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8120229950148366e-05,
+      "loss": 2.0373,
+      "step": 10404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8119880015060582e-05,
+      "loss": 2.0979,
+      "step": 10405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.811953005078375e-05,
+      "loss": 1.724,
+      "step": 10406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8119180057319126e-05,
+      "loss": 1.9514,
+      "step": 10407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.811883003466797e-05,
+      "loss": 1.601,
+      "step": 10408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8118479982831538e-05,
+      "loss": 1.8741,
+      "step": 10409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8118129901811094e-05,
+      "loss": 2.016,
+      "step": 10410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8117779791607892e-05,
+      "loss": 1.9753,
+      "step": 10411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8117429652223192e-05,
+      "loss": 1.4926,
+      "step": 10412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.811707948365825e-05,
+      "loss": 2.2238,
+      "step": 10413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.811672928591433e-05,
+      "loss": 2.018,
+      "step": 10414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8116379058992684e-05,
+      "loss": 1.6728,
+      "step": 10415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8116028802894576e-05,
+      "loss": 1.9051,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8115678517621265e-05,
+      "loss": 1.8485,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.811532820317401e-05,
+      "loss": 1.6218,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8114977859554065e-05,
+      "loss": 1.8672,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8114627486762697e-05,
+      "loss": 1.8575,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8114277084801162e-05,
+      "loss": 1.7461,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8113926653670718e-05,
+      "loss": 1.7774,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.811357619337263e-05,
+      "loss": 1.9418,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8113225703908153e-05,
+      "loss": 1.703,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8112875185278546e-05,
+      "loss": 2.1563,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8112524637485072e-05,
+      "loss": 1.7838,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8112174060528996e-05,
+      "loss": 2.162,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8111823454411568e-05,
+      "loss": 2.5629,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.811147281913405e-05,
+      "loss": 2.0482,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.811112215469771e-05,
+      "loss": 2.0005,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8110771461103805e-05,
+      "loss": 2.1684,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8110420738353592e-05,
+      "loss": 1.8759,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.8110069986448336e-05,
+      "loss": 0.1142,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8109719205389297e-05,
+      "loss": 2.0973,
+      "step": 10434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.810936839517773e-05,
+      "loss": 1.9127,
+      "step": 10435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8109017555814908e-05,
+      "loss": 2.2561,
+      "step": 10436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.810866668730208e-05,
+      "loss": 1.9711,
+      "step": 10437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8108315789640517e-05,
+      "loss": 1.9094,
+      "step": 10438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.375,
+      "learning_rate": 1.810796486283148e-05,
+      "loss": 1.9462,
+      "step": 10439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.810761390687622e-05,
+      "loss": 2.0466,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.8107262921776007e-05,
+      "loss": 1.9162,
+      "step": 10441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.81069119075321e-05,
+      "loss": 1.7526,
+      "step": 10442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8106560864145765e-05,
+      "loss": 1.8774,
+      "step": 10443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.810620979161826e-05,
+      "loss": 1.8087,
+      "step": 10444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8105858689950847e-05,
+      "loss": 1.7964,
+      "step": 10445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.810550755914479e-05,
+      "loss": 1.9333,
+      "step": 10446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.810515639920135e-05,
+      "loss": 1.7795,
+      "step": 10447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8104805210121786e-05,
+      "loss": 1.925,
+      "step": 10448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8104453991907366e-05,
+      "loss": 2.1623,
+      "step": 10449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8104102744559356e-05,
+      "loss": 1.874,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.8103751468079006e-05,
+      "loss": 2.091,
+      "step": 10451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.8103400162467587e-05,
+      "loss": 1.8784,
+      "step": 10452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8103048827726363e-05,
+      "loss": 1.8997,
+      "step": 10453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8102697463856594e-05,
+      "loss": 2.1249,
+      "step": 10454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8102346070859548e-05,
+      "loss": 1.9794,
+      "step": 10455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8101994648736477e-05,
+      "loss": 1.9255,
+      "step": 10456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.8101643197488656e-05,
+      "loss": 1.76,
+      "step": 10457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.8101291717117342e-05,
+      "loss": 2.1664,
+      "step": 10458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.8100940207623802e-05,
+      "loss": 1.4473,
+      "step": 10459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8100588669009296e-05,
+      "loss": 2.2649,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.810023710127509e-05,
+      "loss": 1.7693,
+      "step": 10461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8099885504422447e-05,
+      "loss": 1.909,
+      "step": 10462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.809953387845263e-05,
+      "loss": 1.8025,
+      "step": 10463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8099182223366903e-05,
+      "loss": 1.7378,
+      "step": 10464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8098830539166536e-05,
+      "loss": 1.6746,
+      "step": 10465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8098478825852786e-05,
+      "loss": 1.9005,
+      "step": 10466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.809812708342692e-05,
+      "loss": 1.9488,
+      "step": 10467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8097775311890202e-05,
+      "loss": 1.6387,
+      "step": 10468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8097423511243896e-05,
+      "loss": 2.0623,
+      "step": 10469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.809707168148927e-05,
+      "loss": 2.1537,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.809671982262758e-05,
+      "loss": 1.9575,
+      "step": 10471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8096367934660106e-05,
+      "loss": 1.5847,
+      "step": 10472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.80960160175881e-05,
+      "loss": 1.9789,
+      "step": 10473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.809566407141283e-05,
+      "loss": 1.9945,
+      "step": 10474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8095312096135562e-05,
+      "loss": 2.0039,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8094960091757564e-05,
+      "loss": 1.8767,
+      "step": 10476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8094608058280098e-05,
+      "loss": 1.923,
+      "step": 10477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.809425599570443e-05,
+      "loss": 2.1721,
+      "step": 10478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8093903904031825e-05,
+      "loss": 1.6961,
+      "step": 10479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.809355178326355e-05,
+      "loss": 1.8556,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8093199633400872e-05,
+      "loss": 1.7833,
+      "step": 10481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8092847454445054e-05,
+      "loss": 1.9295,
+      "step": 10482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.809249524639736e-05,
+      "loss": 1.7819,
+      "step": 10483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8092143009259063e-05,
+      "loss": 1.9445,
+      "step": 10484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8091790743031428e-05,
+      "loss": 1.8914,
+      "step": 10485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8091438447715715e-05,
+      "loss": 1.5093,
+      "step": 10486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8091086123313196e-05,
+      "loss": 2.12,
+      "step": 10487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8090733769825135e-05,
+      "loss": 1.9052,
+      "step": 10488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.80903813872528e-05,
+      "loss": 1.7025,
+      "step": 10489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.809002897559746e-05,
+      "loss": 1.9197,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8089676534860376e-05,
+      "loss": 2.1312,
+      "step": 10491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.808932406504282e-05,
+      "loss": 2.2941,
+      "step": 10492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8088971566146054e-05,
+      "loss": 2.092,
+      "step": 10493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.808861903817135e-05,
+      "loss": 1.8534,
+      "step": 10494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.808826648111997e-05,
+      "loss": 1.7712,
+      "step": 10495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.808791389499319e-05,
+      "loss": 2.2146,
+      "step": 10496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.808756127979227e-05,
+      "loss": 1.8703,
+      "step": 10497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.808720863551848e-05,
+      "loss": 2.0679,
+      "step": 10498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8086855962173085e-05,
+      "loss": 1.5971,
+      "step": 10499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.808650325975736e-05,
+      "loss": 1.7832,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8086150528272565e-05,
+      "loss": 1.7111,
+      "step": 10501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8085797767719974e-05,
+      "loss": 1.8031,
+      "step": 10502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8085444978100852e-05,
+      "loss": 2.1898,
+      "step": 10503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8085092159416465e-05,
+      "loss": 1.7502,
+      "step": 10504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8084739311668083e-05,
+      "loss": 1.6436,
+      "step": 10505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.808438643485698e-05,
+      "loss": 2.1851,
+      "step": 10506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8084033528984414e-05,
+      "loss": 1.7728,
+      "step": 10507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8083680594051663e-05,
+      "loss": 1.817,
+      "step": 10508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.808332763005999e-05,
+      "loss": 1.3691,
+      "step": 10509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.808297463701067e-05,
+      "loss": 2.0762,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8082621614904963e-05,
+      "loss": 1.9637,
+      "step": 10511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8082268563744144e-05,
+      "loss": 1.9648,
+      "step": 10512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.808191548352948e-05,
+      "loss": 1.8352,
+      "step": 10513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8081562374262242e-05,
+      "loss": 2.1635,
+      "step": 10514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.80812092359437e-05,
+      "loss": 1.7999,
+      "step": 10515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.808085606857512e-05,
+      "loss": 1.9083,
+      "step": 10516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.8080502872157774e-05,
+      "loss": 2.1451,
+      "step": 10517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8080149646692932e-05,
+      "loss": 1.632,
+      "step": 10518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.807979639218186e-05,
+      "loss": 1.9776,
+      "step": 10519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8079443108625832e-05,
+      "loss": 1.7031,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.807908979602612e-05,
+      "loss": 2.2857,
+      "step": 10521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.807873645438399e-05,
+      "loss": 2.0036,
+      "step": 10522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8078383083700712e-05,
+      "loss": 1.8562,
+      "step": 10523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8078029683977555e-05,
+      "loss": 1.7947,
+      "step": 10524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8077676255215793e-05,
+      "loss": 1.7849,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8077322797416695e-05,
+      "loss": 1.8825,
+      "step": 10526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8076969310581535e-05,
+      "loss": 2.0429,
+      "step": 10527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.807661579471158e-05,
+      "loss": 1.9209,
+      "step": 10528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8076262249808096e-05,
+      "loss": 1.8211,
+      "step": 10529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8075908675872365e-05,
+      "loss": 1.9488,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.807555507290565e-05,
+      "loss": 2.0521,
+      "step": 10531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8075201440909223e-05,
+      "loss": 1.8386,
+      "step": 10532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8074847779884357e-05,
+      "loss": 1.9236,
+      "step": 10533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8074494089832325e-05,
+      "loss": 1.6569,
+      "step": 10534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8074140370754395e-05,
+      "loss": 1.7869,
+      "step": 10535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.807378662265184e-05,
+      "loss": 1.8018,
+      "step": 10536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.807343284552593e-05,
+      "loss": 2.5187,
+      "step": 10537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8073079039377942e-05,
+      "loss": 2.0451,
+      "step": 10538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.807272520420914e-05,
+      "loss": 1.8292,
+      "step": 10539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.80723713400208e-05,
+      "loss": 2.0526,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.80720174468142e-05,
+      "loss": 2.0681,
+      "step": 10541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.80716635245906e-05,
+      "loss": 2.0341,
+      "step": 10542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8071309573351278e-05,
+      "loss": 1.8973,
+      "step": 10543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.807095559309751e-05,
+      "loss": 2.2491,
+      "step": 10544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8070601583830566e-05,
+      "loss": 1.7586,
+      "step": 10545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8070247545551716e-05,
+      "loss": 2.019,
+      "step": 10546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8069893478262238e-05,
+      "loss": 1.7403,
+      "step": 10547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8069539381963396e-05,
+      "loss": 2.0451,
+      "step": 10548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.806918525665647e-05,
+      "loss": 1.8411,
+      "step": 10549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.8068831102342736e-05,
+      "loss": 1.6526,
+      "step": 10550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8068476919023456e-05,
+      "loss": 1.8334,
+      "step": 10551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8068122706699914e-05,
+      "loss": 2.105,
+      "step": 10552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8067768465373375e-05,
+      "loss": 2.0744,
+      "step": 10553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.806741419504512e-05,
+      "loss": 1.9229,
+      "step": 10554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8067059895716416e-05,
+      "loss": 1.8515,
+      "step": 10555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.806670556738854e-05,
+      "loss": 1.8122,
+      "step": 10556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8066351210062768e-05,
+      "loss": 1.9762,
+      "step": 10557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.806599682374037e-05,
+      "loss": 2.0301,
+      "step": 10558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.806564240842262e-05,
+      "loss": 1.6552,
+      "step": 10559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8065287964110792e-05,
+      "loss": 1.9987,
+      "step": 10560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.806493349080616e-05,
+      "loss": 2.0546,
+      "step": 10561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.8064578988510004e-05,
+      "loss": 1.9205,
+      "step": 10562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8064224457223592e-05,
+      "loss": 1.7209,
+      "step": 10563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.80638698969482e-05,
+      "loss": 1.9992,
+      "step": 10564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8063515307685103e-05,
+      "loss": 1.8962,
+      "step": 10565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8063160689435573e-05,
+      "loss": 2.0596,
+      "step": 10566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.806280604220089e-05,
+      "loss": 1.998,
+      "step": 10567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8062451365982324e-05,
+      "loss": 2.4965,
+      "step": 10568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8062096660781153e-05,
+      "loss": 1.8178,
+      "step": 10569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8061741926598654e-05,
+      "loss": 1.5482,
+      "step": 10570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8061387163436095e-05,
+      "loss": 2.0234,
+      "step": 10571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.806103237129476e-05,
+      "loss": 2.0905,
+      "step": 10572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8060677550175914e-05,
+      "loss": 2.2803,
+      "step": 10573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8060322700080846e-05,
+      "loss": 1.977,
+      "step": 10574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8059967821010817e-05,
+      "loss": 1.8852,
+      "step": 10575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.8059612912967112e-05,
+      "loss": 1.9866,
+      "step": 10576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8059257975951007e-05,
+      "loss": 1.7402,
+      "step": 10577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8058903009963776e-05,
+      "loss": 1.7595,
+      "step": 10578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.805854801500669e-05,
+      "loss": 1.7646,
+      "step": 10579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8058192991081035e-05,
+      "loss": 1.4831,
+      "step": 10580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8057837938188078e-05,
+      "loss": 1.9087,
+      "step": 10581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.80574828563291e-05,
+      "loss": 1.7543,
+      "step": 10582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8057127745505377e-05,
+      "loss": 1.9625,
+      "step": 10583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8056772605718183e-05,
+      "loss": 2.0492,
+      "step": 10584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.80564174369688e-05,
+      "loss": 1.8769,
+      "step": 10585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.80560622392585e-05,
+      "loss": 1.8375,
+      "step": 10586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.8055707012588563e-05,
+      "loss": 1.9519,
+      "step": 10587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.8055351756960262e-05,
+      "loss": 1.9125,
+      "step": 10588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.805499647237488e-05,
+      "loss": 1.83,
+      "step": 10589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8054641158833687e-05,
+      "loss": 1.8502,
+      "step": 10590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8054285816337967e-05,
+      "loss": 1.9984,
+      "step": 10591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8053930444888994e-05,
+      "loss": 1.6961,
+      "step": 10592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.805357504448804e-05,
+      "loss": 2.0349,
+      "step": 10593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8053219615136398e-05,
+      "loss": 1.68,
+      "step": 10594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.805286415683533e-05,
+      "loss": 1.6917,
+      "step": 10595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.805250866958612e-05,
+      "loss": 1.773,
+      "step": 10596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8052153153390045e-05,
+      "loss": 1.9256,
+      "step": 10597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8051797608248387e-05,
+      "loss": 1.9762,
+      "step": 10598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.8051442034162417e-05,
+      "loss": 1.9482,
+      "step": 10599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.805108643113342e-05,
+      "loss": 1.8032,
+      "step": 10600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.805073079916267e-05,
+      "loss": 1.984,
+      "step": 10601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8050375138251448e-05,
+      "loss": 1.7913,
+      "step": 10602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.805001944840103e-05,
+      "loss": 1.7287,
+      "step": 10603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.8049663729612693e-05,
+      "loss": 1.821,
+      "step": 10604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8049307981887724e-05,
+      "loss": 2.0265,
+      "step": 10605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.8048952205227394e-05,
+      "loss": 2.2021,
+      "step": 10606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8048596399632985e-05,
+      "loss": 1.9143,
+      "step": 10607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.8048240565105773e-05,
+      "loss": 1.766,
+      "step": 10608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.804788470164704e-05,
+      "loss": 1.9385,
+      "step": 10609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.804752880925807e-05,
+      "loss": 2.1011,
+      "step": 10610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.804717288794013e-05,
+      "loss": 1.963,
+      "step": 10611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8046816937694507e-05,
+      "loss": 2.2,
+      "step": 10612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8046460958522485e-05,
+      "loss": 2.7326,
+      "step": 10613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8046104950425337e-05,
+      "loss": 1.6772,
+      "step": 10614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8045748913404345e-05,
+      "loss": 1.9082,
+      "step": 10615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.804539284746079e-05,
+      "loss": 1.6999,
+      "step": 10616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8045036752595946e-05,
+      "loss": 1.657,
+      "step": 10617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.80446806288111e-05,
+      "loss": 2.2368,
+      "step": 10618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8044324476107528e-05,
+      "loss": 2.1253,
+      "step": 10619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8043968294486513e-05,
+      "loss": 1.7043,
+      "step": 10620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8043612083949334e-05,
+      "loss": 1.8448,
+      "step": 10621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.8043255844497274e-05,
+      "loss": 1.8817,
+      "step": 10622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8042899576131608e-05,
+      "loss": 1.9079,
+      "step": 10623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8042543278853624e-05,
+      "loss": 2.0233,
+      "step": 10624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8042186952664596e-05,
+      "loss": 2.1414,
+      "step": 10625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.375,
+      "learning_rate": 1.804183059756581e-05,
+      "loss": 2.0038,
+      "step": 10626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.804147421355854e-05,
+      "loss": 1.8805,
+      "step": 10627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8041117800644076e-05,
+      "loss": 1.9728,
+      "step": 10628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.8040761358823694e-05,
+      "loss": 2.4819,
+      "step": 10629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.804040488809868e-05,
+      "loss": 1.8667,
+      "step": 10630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8040048388470307e-05,
+      "loss": 2.002,
+      "step": 10631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8039691859939864e-05,
+      "loss": 1.8511,
+      "step": 10632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.8039335302508626e-05,
+      "loss": 1.913,
+      "step": 10633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8038978716177884e-05,
+      "loss": 2.0622,
+      "step": 10634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8038622100948916e-05,
+      "loss": 1.6124,
+      "step": 10635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8038265456823e-05,
+      "loss": 1.7427,
+      "step": 10636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.803790878380142e-05,
+      "loss": 1.9303,
+      "step": 10637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8037552081885457e-05,
+      "loss": 1.9816,
+      "step": 10638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8037195351076397e-05,
+      "loss": 1.9727,
+      "step": 10639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8036838591375517e-05,
+      "loss": 2.0601,
+      "step": 10640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.803648180278411e-05,
+      "loss": 2.032,
+      "step": 10641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.803612498530345e-05,
+      "loss": 1.5524,
+      "step": 10642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8035768138934816e-05,
+      "loss": 1.9079,
+      "step": 10643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.80354112636795e-05,
+      "loss": 1.6231,
+      "step": 10644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8035054359538778e-05,
+      "loss": 1.6633,
+      "step": 10645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.803469742651394e-05,
+      "loss": 2.0203,
+      "step": 10646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8034340464606263e-05,
+      "loss": 2.0249,
+      "step": 10647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.803398347381703e-05,
+      "loss": 1.834,
+      "step": 10648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.803362645414753e-05,
+      "loss": 1.9595,
+      "step": 10649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.803326940559904e-05,
+      "loss": 1.7111,
+      "step": 10650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.803291232817285e-05,
+      "loss": 1.8272,
+      "step": 10651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8032555221870236e-05,
+      "loss": 2.0451,
+      "step": 10652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.8032198086692485e-05,
+      "loss": 1.9811,
+      "step": 10653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8031840922640885e-05,
+      "loss": 1.9401,
+      "step": 10654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.803148372971671e-05,
+      "loss": 1.936,
+      "step": 10655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.803112650792126e-05,
+      "loss": 1.9416,
+      "step": 10656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8030769257255804e-05,
+      "loss": 1.6829,
+      "step": 10657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.803041197772163e-05,
+      "loss": 1.816,
+      "step": 10658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8030054669320027e-05,
+      "loss": 1.8152,
+      "step": 10659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8029697332052277e-05,
+      "loss": 1.977,
+      "step": 10660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.8029339965919663e-05,
+      "loss": 1.7189,
+      "step": 10661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.802898257092347e-05,
+      "loss": 1.7483,
+      "step": 10662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8028625147064988e-05,
+      "loss": 2.0662,
+      "step": 10663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8028267694345492e-05,
+      "loss": 2.0767,
+      "step": 10664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8027910212766272e-05,
+      "loss": 1.8165,
+      "step": 10665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8027552702328615e-05,
+      "loss": 1.863,
+      "step": 10666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8027195163033804e-05,
+      "loss": 1.7481,
+      "step": 10667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8026837594883124e-05,
+      "loss": 1.9204,
+      "step": 10668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.8026479997877864e-05,
+      "loss": 2.1785,
+      "step": 10669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8026122372019307e-05,
+      "loss": 1.9528,
+      "step": 10670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.8025764717308734e-05,
+      "loss": 1.9482,
+      "step": 10671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8025407033747437e-05,
+      "loss": 2.107,
+      "step": 10672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8025049321336702e-05,
+      "loss": 2.0374,
+      "step": 10673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8024691580077808e-05,
+      "loss": 1.831,
+      "step": 10674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8024333809972047e-05,
+      "loss": 1.7219,
+      "step": 10675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.8023976011020705e-05,
+      "loss": 1.7245,
+      "step": 10676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.8023618183225065e-05,
+      "loss": 1.7775,
+      "step": 10677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.802326032658642e-05,
+      "loss": 2.1299,
+      "step": 10678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8022902441106047e-05,
+      "loss": 1.7943,
+      "step": 10679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8022544526785236e-05,
+      "loss": 1.8375,
+      "step": 10680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8022186583625277e-05,
+      "loss": 1.6356,
+      "step": 10681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8021828611627455e-05,
+      "loss": 1.66,
+      "step": 10682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8021470610793058e-05,
+      "loss": 1.637,
+      "step": 10683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8021112581123368e-05,
+      "loss": 1.9092,
+      "step": 10684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.8020754522619677e-05,
+      "loss": 1.6925,
+      "step": 10685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.802039643528327e-05,
+      "loss": 1.9464,
+      "step": 10686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8020038319115435e-05,
+      "loss": 1.9877,
+      "step": 10687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8019680174117456e-05,
+      "loss": 1.7768,
+      "step": 10688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8019322000290626e-05,
+      "loss": 1.9632,
+      "step": 10689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.8018963797636226e-05,
+      "loss": 1.8941,
+      "step": 10690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8018605566155552e-05,
+      "loss": 1.8789,
+      "step": 10691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8018247305849884e-05,
+      "loss": 2.11,
+      "step": 10692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8017889016720517e-05,
+      "loss": 1.7543,
+      "step": 10693
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8017530698768734e-05,
+      "loss": 2.2554,
+      "step": 10694
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8017172351995825e-05,
+      "loss": 1.8938,
+      "step": 10695
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.375,
+      "learning_rate": 1.8016813976403073e-05,
+      "loss": 1.4835,
+      "step": 10696
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8016455571991774e-05,
+      "loss": 1.877,
+      "step": 10697
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.8016097138763215e-05,
+      "loss": 2.3536,
+      "step": 10698
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8015738676718682e-05,
+      "loss": 1.8912,
+      "step": 10699
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.801538018585946e-05,
+      "loss": 2.0165,
+      "step": 10700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8015021666186846e-05,
+      "loss": 2.0645,
+      "step": 10701
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.801466311770212e-05,
+      "loss": 2.1555,
+      "step": 10702
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.801430454040658e-05,
+      "loss": 1.661,
+      "step": 10703
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8013945934301505e-05,
+      "loss": 1.5705,
+      "step": 10704
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.8013587299388197e-05,
+      "loss": 1.9372,
+      "step": 10705
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.8013228635667936e-05,
+      "loss": 1.8545,
+      "step": 10706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.8012869943142012e-05,
+      "loss": 1.6062,
+      "step": 10707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.8012511221811712e-05,
+      "loss": 1.7865,
+      "step": 10708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8012152471678332e-05,
+      "loss": 1.9575,
+      "step": 10709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8011793692743162e-05,
+      "loss": 1.7749,
+      "step": 10710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8011434885007482e-05,
+      "loss": 1.6221,
+      "step": 10711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.801107604847259e-05,
+      "loss": 2.0464,
+      "step": 10712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8010717183139775e-05,
+      "loss": 2.3979,
+      "step": 10713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.801035828901033e-05,
+      "loss": 1.8982,
+      "step": 10714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.800999936608554e-05,
+      "loss": 1.9809,
+      "step": 10715
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.8009640414366692e-05,
+      "loss": 1.7197,
+      "step": 10716
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8009281433855085e-05,
+      "loss": 2.1582,
+      "step": 10717
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.8008922424552002e-05,
+      "loss": 1.5546,
+      "step": 10718
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8008563386458743e-05,
+      "loss": 1.9554,
+      "step": 10719
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.8008204319576585e-05,
+      "loss": 1.7407,
+      "step": 10720
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.800784522390683e-05,
+      "loss": 1.7448,
+      "step": 10721
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8007486099450767e-05,
+      "loss": 1.8473,
+      "step": 10722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.8007126946209684e-05,
+      "loss": 1.7458,
+      "step": 10723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8006767764184873e-05,
+      "loss": 1.4653,
+      "step": 10724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.8006408553377622e-05,
+      "loss": 1.6188,
+      "step": 10725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8006049313789228e-05,
+      "loss": 1.9864,
+      "step": 10726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8005690045420982e-05,
+      "loss": 2.1317,
+      "step": 10727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.800533074827417e-05,
+      "loss": 1.9117,
+      "step": 10728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8004971422350094e-05,
+      "loss": 2.0136,
+      "step": 10729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.8004612067650034e-05,
+      "loss": 1.7849,
+      "step": 10730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.8004252684175286e-05,
+      "loss": 2.3615,
+      "step": 10731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.8003893271927144e-05,
+      "loss": 1.8885,
+      "step": 10732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.80035338309069e-05,
+      "loss": 1.8132,
+      "step": 10733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.800317436111584e-05,
+      "loss": 1.9287,
+      "step": 10734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8002814862555265e-05,
+      "loss": 1.8479,
+      "step": 10735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.800245533522646e-05,
+      "loss": 1.7136,
+      "step": 10736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.8002095779130725e-05,
+      "loss": 1.8191,
+      "step": 10737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.8001736194269344e-05,
+      "loss": 2.0464,
+      "step": 10738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.8001376580643617e-05,
+      "loss": 1.8841,
+      "step": 10739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.800101693825483e-05,
+      "loss": 1.9518,
+      "step": 10740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.8000657267104282e-05,
+      "loss": 1.7183,
+      "step": 10741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.8000297567193265e-05,
+      "loss": 1.7952,
+      "step": 10742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7999937838523066e-05,
+      "loss": 1.7085,
+      "step": 10743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.799957808109498e-05,
+      "loss": 1.9012,
+      "step": 10744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.799921829491031e-05,
+      "loss": 1.5244,
+      "step": 10745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.799885847997034e-05,
+      "loss": 1.9388,
+      "step": 10746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7998498636276363e-05,
+      "loss": 2.0743,
+      "step": 10747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7998138763829677e-05,
+      "loss": 2.3235,
+      "step": 10748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.799777886263157e-05,
+      "loss": 1.8263,
+      "step": 10749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.799741893268334e-05,
+      "loss": 1.4432,
+      "step": 10750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7997058973986285e-05,
+      "loss": 1.946,
+      "step": 10751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.799669898654169e-05,
+      "loss": 1.926,
+      "step": 10752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7996338970350858e-05,
+      "loss": 1.8584,
+      "step": 10753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7995978925415074e-05,
+      "loss": 2.523,
+      "step": 10754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.799561885173564e-05,
+      "loss": 2.4133,
+      "step": 10755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7995258749313845e-05,
+      "loss": 1.9724,
+      "step": 10756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7994898618150985e-05,
+      "loss": 1.824,
+      "step": 10757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7994538458248357e-05,
+      "loss": 2.0968,
+      "step": 10758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.799417826960725e-05,
+      "loss": 1.7874,
+      "step": 10759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.799381805222897e-05,
+      "loss": 2.6331,
+      "step": 10760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7993457806114797e-05,
+      "loss": 1.9438,
+      "step": 10761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7993097531266038e-05,
+      "loss": 1.9981,
+      "step": 10762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7992737227683984e-05,
+      "loss": 1.9731,
+      "step": 10763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7992376895369928e-05,
+      "loss": 1.9143,
+      "step": 10764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7992016534325168e-05,
+      "loss": 1.8898,
+      "step": 10765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7991656144550996e-05,
+      "loss": 1.783,
+      "step": 10766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7991295726048715e-05,
+      "loss": 1.9097,
+      "step": 10767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7990935278819614e-05,
+      "loss": 1.7972,
+      "step": 10768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.799057480286499e-05,
+      "loss": 1.737,
+      "step": 10769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7990214298186136e-05,
+      "loss": 1.8477,
+      "step": 10770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7989853764784354e-05,
+      "loss": 2.0842,
+      "step": 10771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.798949320266094e-05,
+      "loss": 2.0521,
+      "step": 10772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7989132611817184e-05,
+      "loss": 2.1022,
+      "step": 10773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7988771992254385e-05,
+      "loss": 2.0359,
+      "step": 10774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7988411343973845e-05,
+      "loss": 1.7504,
+      "step": 10775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.798805066697685e-05,
+      "loss": 1.7358,
+      "step": 10776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.7987689961264702e-05,
+      "loss": 2.0357,
+      "step": 10777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7987329226838702e-05,
+      "loss": 2.1325,
+      "step": 10778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.798696846370014e-05,
+      "loss": 1.9038,
+      "step": 10779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7986607671850312e-05,
+      "loss": 2.288,
+      "step": 10780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.7986246851290524e-05,
+      "loss": 1.9075,
+      "step": 10781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7985886002022063e-05,
+      "loss": 2.1725,
+      "step": 10782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.798552512404623e-05,
+      "loss": 1.7056,
+      "step": 10783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7985164217364328e-05,
+      "loss": 2.1707,
+      "step": 10784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7984803281977647e-05,
+      "loss": 2.0502,
+      "step": 10785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7984442317887482e-05,
+      "loss": 1.8311,
+      "step": 10786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.798408132509514e-05,
+      "loss": 2.0437,
+      "step": 10787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7983720303601914e-05,
+      "loss": 2.0316,
+      "step": 10788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.79833592534091e-05,
+      "loss": 2.0681,
+      "step": 10789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7982998174517994e-05,
+      "loss": 1.7246,
+      "step": 10790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7982637066929905e-05,
+      "loss": 1.9294,
+      "step": 10791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7982275930646118e-05,
+      "loss": 2.0404,
+      "step": 10792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7981914765667938e-05,
+      "loss": 1.7293,
+      "step": 10793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7981553571996662e-05,
+      "loss": 2.0766,
+      "step": 10794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.798119234963359e-05,
+      "loss": 1.6063,
+      "step": 10795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7980831098580016e-05,
+      "loss": 1.5973,
+      "step": 10796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7980469818837243e-05,
+      "loss": 2.1703,
+      "step": 10797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.798010851040657e-05,
+      "loss": 1.8003,
+      "step": 10798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7979747173289292e-05,
+      "loss": 1.9406,
+      "step": 10799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7979385807486712e-05,
+      "loss": 2.0689,
+      "step": 10800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7979024413000127e-05,
+      "loss": 1.71,
+      "step": 10801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7978662989830834e-05,
+      "loss": 1.7262,
+      "step": 10802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.797830153798014e-05,
+      "loss": 1.9655,
+      "step": 10803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.797794005744933e-05,
+      "loss": 1.807,
+      "step": 10804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.797757854823972e-05,
+      "loss": 2.0048,
+      "step": 10805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7977217010352597e-05,
+      "loss": 1.8619,
+      "step": 10806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7976855443789266e-05,
+      "loss": 1.9951,
+      "step": 10807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7976493848551027e-05,
+      "loss": 1.7465,
+      "step": 10808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.797613222463918e-05,
+      "loss": 1.7992,
+      "step": 10809
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7975770572055024e-05,
+      "loss": 2.1903,
+      "step": 10810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7975408890799857e-05,
+      "loss": 1.9776,
+      "step": 10811
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7975047180874984e-05,
+      "loss": 2.2912,
+      "step": 10812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7974685442281697e-05,
+      "loss": 1.6054,
+      "step": 10813
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.797432367502131e-05,
+      "loss": 2.4831,
+      "step": 10814
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7973961879095107e-05,
+      "loss": 1.6479,
+      "step": 10815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7973600054504402e-05,
+      "loss": 2.0479,
+      "step": 10816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.7973238201250488e-05,
+      "loss": 2.4787,
+      "step": 10817
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.797287631933467e-05,
+      "loss": 1.7534,
+      "step": 10818
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7972514408758245e-05,
+      "loss": 1.9934,
+      "step": 10819
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.797215246952252e-05,
+      "loss": 1.9585,
+      "step": 10820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.797179050162879e-05,
+      "loss": 2.0369,
+      "step": 10821
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.797142850507836e-05,
+      "loss": 1.8394,
+      "step": 10822
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7971066479872525e-05,
+      "loss": 1.7241,
+      "step": 10823
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7970704426012597e-05,
+      "loss": 1.908,
+      "step": 10824
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7970342343499868e-05,
+      "loss": 1.861,
+      "step": 10825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7969980232335643e-05,
+      "loss": 1.9652,
+      "step": 10826
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7969618092521223e-05,
+      "loss": 1.7741,
+      "step": 10827
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7969255924057917e-05,
+      "loss": 1.8956,
+      "step": 10828
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7968893726947014e-05,
+      "loss": 2.1244,
+      "step": 10829
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7968531501189822e-05,
+      "loss": 1.7658,
+      "step": 10830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.796816924678765e-05,
+      "loss": 2.1817,
+      "step": 10831
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.796780696374179e-05,
+      "loss": 2.0989,
+      "step": 10832
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7967444652053546e-05,
+      "loss": 1.5503,
+      "step": 10833
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.796708231172423e-05,
+      "loss": 2.0082,
+      "step": 10834
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.796671994275513e-05,
+      "loss": 1.9933,
+      "step": 10835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.796635754514756e-05,
+      "loss": 1.9197,
+      "step": 10836
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.7965995118902816e-05,
+      "loss": 2.165,
+      "step": 10837
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7965632664022206e-05,
+      "loss": 1.8762,
+      "step": 10838
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7965270180507028e-05,
+      "loss": 1.8725,
+      "step": 10839
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7964907668358588e-05,
+      "loss": 2.4509,
+      "step": 10840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7964545127578192e-05,
+      "loss": 1.9392,
+      "step": 10841
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7964182558167137e-05,
+      "loss": 1.7574,
+      "step": 10842
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7963819960126727e-05,
+      "loss": 1.9587,
+      "step": 10843
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.796345733345827e-05,
+      "loss": 2.0682,
+      "step": 10844
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.796309467816307e-05,
+      "loss": 2.0111,
+      "step": 10845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7962731994242423e-05,
+      "loss": 1.5966,
+      "step": 10846
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7962369281697637e-05,
+      "loss": 2.3968,
+      "step": 10847
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.796200654053002e-05,
+      "loss": 1.5169,
+      "step": 10848
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7961643770740875e-05,
+      "loss": 1.8175,
+      "step": 10849
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7961280972331498e-05,
+      "loss": 1.8565,
+      "step": 10850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7960918145303202e-05,
+      "loss": 1.6187,
+      "step": 10851
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7960555289657288e-05,
+      "loss": 1.7308,
+      "step": 10852
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7960192405395057e-05,
+      "loss": 1.8017,
+      "step": 10853
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7959829492517817e-05,
+      "loss": 1.9136,
+      "step": 10854
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7959466551026876e-05,
+      "loss": 1.7216,
+      "step": 10855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7959103580923534e-05,
+      "loss": 1.7272,
+      "step": 10856
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7958740582209097e-05,
+      "loss": 1.9799,
+      "step": 10857
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7958377554884868e-05,
+      "loss": 1.7502,
+      "step": 10858
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7958014498952155e-05,
+      "loss": 1.8611,
+      "step": 10859
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.795765141441226e-05,
+      "loss": 1.6924,
+      "step": 10860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7957288301266494e-05,
+      "loss": 1.9953,
+      "step": 10861
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7956925159516154e-05,
+      "loss": 2.0563,
+      "step": 10862
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7956561989162552e-05,
+      "loss": 1.7319,
+      "step": 10863
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.795619879020699e-05,
+      "loss": 1.9347,
+      "step": 10864
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7955835562650777e-05,
+      "loss": 2.1765,
+      "step": 10865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7955472306495217e-05,
+      "loss": 1.8633,
+      "step": 10866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7955109021741616e-05,
+      "loss": 2.6025,
+      "step": 10867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7954745708391278e-05,
+      "loss": 1.9657,
+      "step": 10868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7954382366445507e-05,
+      "loss": 2.0246,
+      "step": 10869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.795401899590562e-05,
+      "loss": 1.6629,
+      "step": 10870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7953655596772907e-05,
+      "loss": 2.0117,
+      "step": 10871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7953292169048692e-05,
+      "loss": 1.9189,
+      "step": 10872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.7952928712734266e-05,
+      "loss": 2.9633,
+      "step": 10873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7952565227830944e-05,
+      "loss": 1.6649,
+      "step": 10874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7952201714340032e-05,
+      "loss": 1.884,
+      "step": 10875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7951838172262836e-05,
+      "loss": 1.9657,
+      "step": 10876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.795147460160066e-05,
+      "loss": 1.765,
+      "step": 10877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.7951111002354814e-05,
+      "loss": 1.7453,
+      "step": 10878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7950747374526605e-05,
+      "loss": 1.7434,
+      "step": 10879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.795038371811734e-05,
+      "loss": 1.5575,
+      "step": 10880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.795002003312832e-05,
+      "loss": 1.8479,
+      "step": 10881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7949656319560867e-05,
+      "loss": 1.9438,
+      "step": 10882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7949292577416274e-05,
+      "loss": 2.2127,
+      "step": 10883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7948928806695853e-05,
+      "loss": 1.9337,
+      "step": 10884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7948565007400917e-05,
+      "loss": 2.3175,
+      "step": 10885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7948201179532768e-05,
+      "loss": 1.8191,
+      "step": 10886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7947837323092717e-05,
+      "loss": 1.8389,
+      "step": 10887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.794747343808207e-05,
+      "loss": 2.039,
+      "step": 10888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7947109524502135e-05,
+      "loss": 1.9803,
+      "step": 10889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7946745582354218e-05,
+      "loss": 2.3886,
+      "step": 10890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.794638161163963e-05,
+      "loss": 1.8786,
+      "step": 10891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7946017612359685e-05,
+      "loss": 1.8956,
+      "step": 10892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.794565358451568e-05,
+      "loss": 2.2239,
+      "step": 10893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.794528952810893e-05,
+      "loss": 1.7981,
+      "step": 10894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7944925443140746e-05,
+      "loss": 1.8306,
+      "step": 10895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7944561329612434e-05,
+      "loss": 2.0479,
+      "step": 10896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.79441971875253e-05,
+      "loss": 1.9692,
+      "step": 10897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7943833016880658e-05,
+      "loss": 2.0005,
+      "step": 10898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7943468817679814e-05,
+      "loss": 1.888,
+      "step": 10899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.794310458992408e-05,
+      "loss": 1.9962,
+      "step": 10900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.794274033361476e-05,
+      "loss": 1.9144,
+      "step": 10901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.794237604875317e-05,
+      "loss": 1.943,
+      "step": 10902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7942011735340617e-05,
+      "loss": 2.0581,
+      "step": 10903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7941647393378407e-05,
+      "loss": 1.8978,
+      "step": 10904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7941283022867854e-05,
+      "loss": 1.9268,
+      "step": 10905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7940918623810267e-05,
+      "loss": 1.8701,
+      "step": 10906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7940554196206956e-05,
+      "loss": 1.9422,
+      "step": 10907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.794018974005923e-05,
+      "loss": 1.8695,
+      "step": 10908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7939825255368398e-05,
+      "loss": 1.8711,
+      "step": 10909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7939460742135778e-05,
+      "loss": 1.6826,
+      "step": 10910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.7939096200362667e-05,
+      "loss": 2.3286,
+      "step": 10911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7938731630050387e-05,
+      "loss": 2.2027,
+      "step": 10912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7938367031200242e-05,
+      "loss": 1.6266,
+      "step": 10913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.7938002403813544e-05,
+      "loss": 1.8882,
+      "step": 10914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7937637747891607e-05,
+      "loss": 1.6237,
+      "step": 10915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.793727306343574e-05,
+      "loss": 1.9132,
+      "step": 10916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7936908350447254e-05,
+      "loss": 1.7057,
+      "step": 10917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7936543608927454e-05,
+      "loss": 1.8873,
+      "step": 10918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.793617883887766e-05,
+      "loss": 1.6412,
+      "step": 10919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7935814040299182e-05,
+      "loss": 1.7164,
+      "step": 10920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.793544921319333e-05,
+      "loss": 1.6845,
+      "step": 10921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7935084357561413e-05,
+      "loss": 1.7442,
+      "step": 10922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.375,
+      "learning_rate": 1.793471947340474e-05,
+      "loss": 1.7658,
+      "step": 10923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.793435456072463e-05,
+      "loss": 1.8441,
+      "step": 10924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7933989619522396e-05,
+      "loss": 1.8141,
+      "step": 10925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.793362464979934e-05,
+      "loss": 1.767,
+      "step": 10926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7933259651556782e-05,
+      "loss": 1.8999,
+      "step": 10927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.7932894624796035e-05,
+      "loss": 1.7207,
+      "step": 10928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7932529569518402e-05,
+      "loss": 1.9838,
+      "step": 10929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.79321644857252e-05,
+      "loss": 1.6699,
+      "step": 10930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.793179937341775e-05,
+      "loss": 2.0839,
+      "step": 10931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.793143423259735e-05,
+      "loss": 1.9036,
+      "step": 10932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7931069063265325e-05,
+      "loss": 1.9571,
+      "step": 10933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.793070386542298e-05,
+      "loss": 1.9944,
+      "step": 10934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.793033863907163e-05,
+      "loss": 2.3153,
+      "step": 10935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.792997338421259e-05,
+      "loss": 1.727,
+      "step": 10936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.792960810084717e-05,
+      "loss": 1.7038,
+      "step": 10937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.7929242788976685e-05,
+      "loss": 1.572,
+      "step": 10938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7928877448602443e-05,
+      "loss": 1.9771,
+      "step": 10939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7928512079725766e-05,
+      "loss": 1.6178,
+      "step": 10940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.792814668234796e-05,
+      "loss": 1.9757,
+      "step": 10941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7927781256470344e-05,
+      "loss": 1.8648,
+      "step": 10942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.792741580209423e-05,
+      "loss": 1.6692,
+      "step": 10943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.792705031922093e-05,
+      "loss": 1.8464,
+      "step": 10944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7926684807851756e-05,
+      "loss": 1.9628,
+      "step": 10945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.792631926798803e-05,
+      "loss": 1.8864,
+      "step": 10946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7925953699631056e-05,
+      "loss": 1.7339,
+      "step": 10947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7925588102782156e-05,
+      "loss": 1.9134,
+      "step": 10948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7925222477442643e-05,
+      "loss": 2.0353,
+      "step": 10949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7924856823613823e-05,
+      "loss": 1.8484,
+      "step": 10950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.792449114129702e-05,
+      "loss": 1.7175,
+      "step": 10951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.792412543049355e-05,
+      "loss": 1.7323,
+      "step": 10952
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7923759691204717e-05,
+      "loss": 1.7322,
+      "step": 10953
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7923393923431846e-05,
+      "loss": 1.9716,
+      "step": 10954
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.792302812717624e-05,
+      "loss": 1.9086,
+      "step": 10955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7922662302439233e-05,
+      "loss": 2.0033,
+      "step": 10956
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.792229644922212e-05,
+      "loss": 2.041,
+      "step": 10957
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7921930567526228e-05,
+      "loss": 1.9286,
+      "step": 10958
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7921564657352868e-05,
+      "loss": 2.0602,
+      "step": 10959
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7921198718703358e-05,
+      "loss": 1.9205,
+      "step": 10960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.792083275157901e-05,
+      "loss": 1.8887,
+      "step": 10961
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7920466755981143e-05,
+      "loss": 1.9291,
+      "step": 10962
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.792010073191107e-05,
+      "loss": 1.6985,
+      "step": 10963
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.791973467937011e-05,
+      "loss": 1.8225,
+      "step": 10964
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7919368598359574e-05,
+      "loss": 2.1694,
+      "step": 10965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7919002488880784e-05,
+      "loss": 1.6679,
+      "step": 10966
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.791863635093505e-05,
+      "loss": 1.964,
+      "step": 10967
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.791827018452369e-05,
+      "loss": 1.6656,
+      "step": 10968
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7917903989648025e-05,
+      "loss": 2.0783,
+      "step": 10969
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7917537766309366e-05,
+      "loss": 1.9464,
+      "step": 10970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7917171514509035e-05,
+      "loss": 2.1309,
+      "step": 10971
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.791680523424834e-05,
+      "loss": 1.9023,
+      "step": 10972
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7916438925528603e-05,
+      "loss": 1.6542,
+      "step": 10973
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.7916072588351138e-05,
+      "loss": 1.9479,
+      "step": 10974
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.791570622271727e-05,
+      "loss": 2.0477,
+      "step": 10975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7915339828628305e-05,
+      "loss": 1.9631,
+      "step": 10976
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.791497340608557e-05,
+      "loss": 1.9326,
+      "step": 10977
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.791460695509037e-05,
+      "loss": 1.8631,
+      "step": 10978
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7914240475644035e-05,
+      "loss": 2.0602,
+      "step": 10979
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7913873967747876e-05,
+      "loss": 1.6762,
+      "step": 10980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7913507431403214e-05,
+      "loss": 2.0854,
+      "step": 10981
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.791314086661136e-05,
+      "loss": 1.7019,
+      "step": 10982
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7912774273373637e-05,
+      "loss": 1.7398,
+      "step": 10983
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.7912407651691365e-05,
+      "loss": 1.9551,
+      "step": 10984
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.791204100156586e-05,
+      "loss": 1.6941,
+      "step": 10985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.7911674322998432e-05,
+      "loss": 2.4138,
+      "step": 10986
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.791130761599041e-05,
+      "loss": 1.5706,
+      "step": 10987
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.791094088054311e-05,
+      "loss": 1.9523,
+      "step": 10988
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7910574116657846e-05,
+      "loss": 2.1023,
+      "step": 10989
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7910207324335938e-05,
+      "loss": 1.956,
+      "step": 10990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7909840503578707e-05,
+      "loss": 2.1774,
+      "step": 10991
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.790947365438747e-05,
+      "loss": 1.6202,
+      "step": 10992
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7909106776763545e-05,
+      "loss": 1.7581,
+      "step": 10993
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.790873987070825e-05,
+      "loss": 1.8866,
+      "step": 10994
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7908372936222907e-05,
+      "loss": 1.7475,
+      "step": 10995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7908005973308835e-05,
+      "loss": 1.9261,
+      "step": 10996
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.790763898196735e-05,
+      "loss": 1.9591,
+      "step": 10997
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7907271962199776e-05,
+      "loss": 1.8934,
+      "step": 10998
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7906904914007426e-05,
+      "loss": 2.039,
+      "step": 10999
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7906537837391622e-05,
+      "loss": 1.672,
+      "step": 11000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7906170732353688e-05,
+      "loss": 1.9161,
+      "step": 11001
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.790580359889494e-05,
+      "loss": 1.9143,
+      "step": 11002
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.790543643701669e-05,
+      "loss": 2.0715,
+      "step": 11003
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.790506924672027e-05,
+      "loss": 1.91,
+      "step": 11004
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7904702028007e-05,
+      "loss": 1.6798,
+      "step": 11005
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.790433478087819e-05,
+      "loss": 1.6516,
+      "step": 11006
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.790396750533517e-05,
+      "loss": 2.0449,
+      "step": 11007
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7903600201379253e-05,
+      "loss": 1.8594,
+      "step": 11008
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7903232869011767e-05,
+      "loss": 1.7553,
+      "step": 11009
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.790286550823402e-05,
+      "loss": 1.9828,
+      "step": 11010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7902498119047348e-05,
+      "loss": 2.0626,
+      "step": 11011
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.790213070145306e-05,
+      "loss": 2.1217,
+      "step": 11012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7901763255452483e-05,
+      "loss": 1.8189,
+      "step": 11013
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7901395781046937e-05,
+      "loss": 1.7167,
+      "step": 11014
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7901028278237736e-05,
+      "loss": 1.6322,
+      "step": 11015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.790066074702621e-05,
+      "loss": 1.9142,
+      "step": 11016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.7900293187413676e-05,
+      "loss": 1.8516,
+      "step": 11017
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.7899925599401462e-05,
+      "loss": 2.1392,
+      "step": 11018
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.789955798299088e-05,
+      "loss": 2.0917,
+      "step": 11019
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7899190338183253e-05,
+      "loss": 1.8706,
+      "step": 11020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7898822664979908e-05,
+      "loss": 1.7403,
+      "step": 11021
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.789845496338216e-05,
+      "loss": 2.1349,
+      "step": 11022
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7898087233391338e-05,
+      "loss": 1.718,
+      "step": 11023
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.789771947500876e-05,
+      "loss": 1.9472,
+      "step": 11024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7897351688235745e-05,
+      "loss": 2.01,
+      "step": 11025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7896983873073622e-05,
+      "loss": 1.7939,
+      "step": 11026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7896616029523707e-05,
+      "loss": 1.8951,
+      "step": 11027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7896248157587327e-05,
+      "loss": 1.8705,
+      "step": 11028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.78958802572658e-05,
+      "loss": 2.039,
+      "step": 11029
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7895512328560453e-05,
+      "loss": 2.2893,
+      "step": 11030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.7895144371472606e-05,
+      "loss": 1.8986,
+      "step": 11031
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.789477638600358e-05,
+      "loss": 1.6787,
+      "step": 11032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7894408372154704e-05,
+      "loss": 1.665,
+      "step": 11033
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7894040329927293e-05,
+      "loss": 1.9939,
+      "step": 11034
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7893672259322675e-05,
+      "loss": 1.7103,
+      "step": 11035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7893304160342173e-05,
+      "loss": 1.9411,
+      "step": 11036
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.789293603298711e-05,
+      "loss": 1.9149,
+      "step": 11037
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7892567877258805e-05,
+      "loss": 2.0742,
+      "step": 11038
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.7892199693158586e-05,
+      "loss": 1.7412,
+      "step": 11039
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7891831480687777e-05,
+      "loss": 1.8405,
+      "step": 11040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.78914632398477e-05,
+      "loss": 1.7623,
+      "step": 11041
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.789109497063968e-05,
+      "loss": 2.0312,
+      "step": 11042
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.789072667306504e-05,
+      "loss": 1.9837,
+      "step": 11043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7890358347125102e-05,
+      "loss": 1.5277,
+      "step": 11044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.788998999282119e-05,
+      "loss": 3.0812,
+      "step": 11045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7889621610154634e-05,
+      "loss": 2.0472,
+      "step": 11046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.788925319912675e-05,
+      "loss": 2.1284,
+      "step": 11047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.788888475973887e-05,
+      "loss": 1.6846,
+      "step": 11048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.788851629199231e-05,
+      "loss": 2.3271,
+      "step": 11049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7888147795888408e-05,
+      "loss": 1.8294,
+      "step": 11050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7887779271428477e-05,
+      "loss": 1.8353,
+      "step": 11051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.788741071861384e-05,
+      "loss": 2.0257,
+      "step": 11052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7887042137445832e-05,
+      "loss": 2.0433,
+      "step": 11053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.788667352792577e-05,
+      "loss": 2.0714,
+      "step": 11054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7886304890054983e-05,
+      "loss": 2.0015,
+      "step": 11055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7885936223834795e-05,
+      "loss": 1.6512,
+      "step": 11056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.788556752926653e-05,
+      "loss": 1.7306,
+      "step": 11057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7885198806351517e-05,
+      "loss": 2.0387,
+      "step": 11058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7884830055091074e-05,
+      "loss": 1.9587,
+      "step": 11059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.7884461275486534e-05,
+      "loss": 2.2043,
+      "step": 11060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7884092467539224e-05,
+      "loss": 2.3933,
+      "step": 11061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7883723631250466e-05,
+      "loss": 1.5962,
+      "step": 11062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7883354766621582e-05,
+      "loss": 2.0567,
+      "step": 11063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7882985873653906e-05,
+      "loss": 1.8435,
+      "step": 11064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7882616952348756e-05,
+      "loss": 1.8522,
+      "step": 11065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7882248002707464e-05,
+      "loss": 2.0132,
+      "step": 11066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.788187902473135e-05,
+      "loss": 1.8217,
+      "step": 11067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7881510018421752e-05,
+      "loss": 1.8867,
+      "step": 11068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7881140983779986e-05,
+      "loss": 1.5945,
+      "step": 11069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7880771920807386e-05,
+      "loss": 1.8974,
+      "step": 11070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.788040282950527e-05,
+      "loss": 1.6214,
+      "step": 11071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.788003370987497e-05,
+      "loss": 1.9904,
+      "step": 11072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7879664561917814e-05,
+      "loss": 1.847,
+      "step": 11073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7879295385635127e-05,
+      "loss": 1.9492,
+      "step": 11074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7878926181028234e-05,
+      "loss": 1.892,
+      "step": 11075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.787855694809847e-05,
+      "loss": 1.9227,
+      "step": 11076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7878187686847154e-05,
+      "loss": 1.8013,
+      "step": 11077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7877818397275616e-05,
+      "loss": 1.7132,
+      "step": 11078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.787744907938518e-05,
+      "loss": 1.9071,
+      "step": 11079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7877079733177185e-05,
+      "loss": 2.0307,
+      "step": 11080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7876710358652945e-05,
+      "loss": 1.7492,
+      "step": 11081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.78763409558138e-05,
+      "loss": 1.8946,
+      "step": 11082
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7875971524661068e-05,
+      "loss": 1.7022,
+      "step": 11083
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7875602065196083e-05,
+      "loss": 1.9563,
+      "step": 11084
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.787523257742017e-05,
+      "loss": 1.6011,
+      "step": 11085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7874863061334658e-05,
+      "loss": 1.7188,
+      "step": 11086
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7874493516940878e-05,
+      "loss": 1.9901,
+      "step": 11087
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.787412394424015e-05,
+      "loss": 1.7752,
+      "step": 11088
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7873754343233812e-05,
+      "loss": 1.8585,
+      "step": 11089
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.787338471392319e-05,
+      "loss": 1.5958,
+      "step": 11090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7873015056309616e-05,
+      "loss": 1.8923,
+      "step": 11091
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.787264537039441e-05,
+      "loss": 2.0605,
+      "step": 11092
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7872275656178904e-05,
+      "loss": 2.0066,
+      "step": 11093
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7871905913664433e-05,
+      "loss": 1.9006,
+      "step": 11094
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.787153614285232e-05,
+      "loss": 1.9513,
+      "step": 11095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7871166343743896e-05,
+      "loss": 1.7712,
+      "step": 11096
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.787079651634049e-05,
+      "loss": 1.9055,
+      "step": 11097
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7870426660643434e-05,
+      "loss": 2.0368,
+      "step": 11098
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7870056776654054e-05,
+      "loss": 2.0711,
+      "step": 11099
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.786968686437368e-05,
+      "loss": 1.9389,
+      "step": 11100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7869316923803642e-05,
+      "loss": 1.803,
+      "step": 11101
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7868946954945277e-05,
+      "loss": 1.5606,
+      "step": 11102
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.78685769577999e-05,
+      "loss": 1.8446,
+      "step": 11103
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7868206932368856e-05,
+      "loss": 1.8032,
+      "step": 11104
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7867836878653467e-05,
+      "loss": 1.6819,
+      "step": 11105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7867466796655065e-05,
+      "loss": 1.605,
+      "step": 11106
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.786709668637498e-05,
+      "loss": 1.8203,
+      "step": 11107
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7866726547814544e-05,
+      "loss": 1.9213,
+      "step": 11108
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7866356380975086e-05,
+      "loss": 2.0076,
+      "step": 11109
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7865986185857938e-05,
+      "loss": 1.827,
+      "step": 11110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7865615962464427e-05,
+      "loss": 1.8501,
+      "step": 11111
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.786524571079589e-05,
+      "loss": 1.6968,
+      "step": 11112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7864875430853653e-05,
+      "loss": 1.6973,
+      "step": 11113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.786450512263905e-05,
+      "loss": 2.0478,
+      "step": 11114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.786413478615341e-05,
+      "loss": 2.0936,
+      "step": 11115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7863764421398063e-05,
+      "loss": 1.8866,
+      "step": 11116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.786339402837435e-05,
+      "loss": 1.8612,
+      "step": 11117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7863023607083588e-05,
+      "loss": 1.7132,
+      "step": 11118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7862653157527117e-05,
+      "loss": 2.0069,
+      "step": 11119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7862282679706272e-05,
+      "loss": 1.8131,
+      "step": 11120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7861912173622373e-05,
+      "loss": 2.1708,
+      "step": 11121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7861541639276762e-05,
+      "loss": 1.9992,
+      "step": 11122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7861171076670768e-05,
+      "loss": 2.0047,
+      "step": 11123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.786080048580573e-05,
+      "loss": 1.9679,
+      "step": 11124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.786042986668296e-05,
+      "loss": 2.1919,
+      "step": 11125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.7860059219303813e-05,
+      "loss": 1.9125,
+      "step": 11126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.785968854366961e-05,
+      "loss": 1.5974,
+      "step": 11127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7859317839781685e-05,
+      "loss": 1.8517,
+      "step": 11128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.785894710764137e-05,
+      "loss": 1.8045,
+      "step": 11129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.785857634725e-05,
+      "loss": 1.9108,
+      "step": 11130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7858205558608906e-05,
+      "loss": 1.5911,
+      "step": 11131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7857834741719423e-05,
+      "loss": 1.8386,
+      "step": 11132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.785746389658288e-05,
+      "loss": 1.8812,
+      "step": 11133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7857093023200615e-05,
+      "loss": 1.7443,
+      "step": 11134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7856722121573956e-05,
+      "loss": 1.5262,
+      "step": 11135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.785635119170424e-05,
+      "loss": 1.8703,
+      "step": 11136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7855980233592796e-05,
+      "loss": 1.9488,
+      "step": 11137
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.785560924724097e-05,
+      "loss": 1.8897,
+      "step": 11138
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.785523823265008e-05,
+      "loss": 1.9627,
+      "step": 11139
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7854867189821466e-05,
+      "loss": 1.9728,
+      "step": 11140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7854496118756464e-05,
+      "loss": 1.949,
+      "step": 11141
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.785412501945641e-05,
+      "loss": 1.8515,
+      "step": 11142
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7853753891922624e-05,
+      "loss": 1.6037,
+      "step": 11143
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7853382736156458e-05,
+      "loss": 1.8269,
+      "step": 11144
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7853011552159233e-05,
+      "loss": 2.1356,
+      "step": 11145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7852640339932294e-05,
+      "loss": 1.8406,
+      "step": 11146
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7852269099476966e-05,
+      "loss": 1.8239,
+      "step": 11147
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.785189783079459e-05,
+      "loss": 1.677,
+      "step": 11148
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7851526533886498e-05,
+      "loss": 1.8783,
+      "step": 11149
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7851155208754024e-05,
+      "loss": 1.8527,
+      "step": 11150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7850783855398504e-05,
+      "loss": 1.7625,
+      "step": 11151
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7850412473821276e-05,
+      "loss": 2.1628,
+      "step": 11152
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.785004106402367e-05,
+      "loss": 1.9893,
+      "step": 11153
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.784966962600702e-05,
+      "loss": 1.7624,
+      "step": 11154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7849298159772664e-05,
+      "loss": 1.8912,
+      "step": 11155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.784892666532194e-05,
+      "loss": 2.0872,
+      "step": 11156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7848555142656182e-05,
+      "loss": 1.9951,
+      "step": 11157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.784818359177672e-05,
+      "loss": 2.3899,
+      "step": 11158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7847812012684896e-05,
+      "loss": 1.9439,
+      "step": 11159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7847440405382044e-05,
+      "loss": 1.8488,
+      "step": 11160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.78470687698695e-05,
+      "loss": 1.6986,
+      "step": 11161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7846697106148602e-05,
+      "loss": 1.7097,
+      "step": 11162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.784632541422068e-05,
+      "loss": 1.5458,
+      "step": 11163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7845953694087074e-05,
+      "loss": 1.8272,
+      "step": 11164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.784558194574912e-05,
+      "loss": 1.7578,
+      "step": 11165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7845210169208154e-05,
+      "loss": 1.8368,
+      "step": 11166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.7844838364465517e-05,
+      "loss": 2.238,
+      "step": 11167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7844466531522534e-05,
+      "loss": 1.7282,
+      "step": 11168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7844094670380558e-05,
+      "loss": 1.8615,
+      "step": 11169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.784372278104091e-05,
+      "loss": 1.6166,
+      "step": 11170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7843350863504932e-05,
+      "loss": 1.8265,
+      "step": 11171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7842978917773966e-05,
+      "loss": 1.7056,
+      "step": 11172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.784260694384935e-05,
+      "loss": 1.7999,
+      "step": 11173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.784223494173241e-05,
+      "loss": 1.9125,
+      "step": 11174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.784186291142449e-05,
+      "loss": 1.8589,
+      "step": 11175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7841490852926932e-05,
+      "loss": 1.8509,
+      "step": 11176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7841118766241065e-05,
+      "loss": 2.328,
+      "step": 11177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7840746651368234e-05,
+      "loss": 1.7079,
+      "step": 11178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.784037450830977e-05,
+      "loss": 2.0162,
+      "step": 11179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7840002337067014e-05,
+      "loss": 1.8961,
+      "step": 11180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7839630137641305e-05,
+      "loss": 1.9293,
+      "step": 11181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.783925791003398e-05,
+      "loss": 1.8915,
+      "step": 11182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7838885654246377e-05,
+      "loss": 1.7986,
+      "step": 11183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7838513370279834e-05,
+      "loss": 1.8497,
+      "step": 11184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7838141058135686e-05,
+      "loss": 1.7505,
+      "step": 11185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7837768717815276e-05,
+      "loss": 2.1018,
+      "step": 11186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7837396349319944e-05,
+      "loss": 1.6476,
+      "step": 11187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.783702395265102e-05,
+      "loss": 1.8754,
+      "step": 11188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7836651527809852e-05,
+      "loss": 2.2759,
+      "step": 11189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7836279074797776e-05,
+      "loss": 1.7304,
+      "step": 11190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.783590659361613e-05,
+      "loss": 1.7648,
+      "step": 11191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.783553408426625e-05,
+      "loss": 1.6179,
+      "step": 11192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.783516154674948e-05,
+      "loss": 1.571,
+      "step": 11193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7834788981067154e-05,
+      "loss": 1.837,
+      "step": 11194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7834416387220617e-05,
+      "loss": 1.815,
+      "step": 11195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7834043765211208e-05,
+      "loss": 1.8802,
+      "step": 11196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7833671115040263e-05,
+      "loss": 1.3657,
+      "step": 11197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7833298436709122e-05,
+      "loss": 1.9897,
+      "step": 11198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7832925730219126e-05,
+      "loss": 2.2184,
+      "step": 11199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7832552995571614e-05,
+      "loss": 1.6135,
+      "step": 11200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7832180232767927e-05,
+      "loss": 1.5345,
+      "step": 11201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7831807441809407e-05,
+      "loss": 1.77,
+      "step": 11202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7831434622697386e-05,
+      "loss": 1.9017,
+      "step": 11203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7831061775433214e-05,
+      "loss": 1.8624,
+      "step": 11204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.783068890001822e-05,
+      "loss": 1.9914,
+      "step": 11205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.375,
+      "learning_rate": 1.783031599645376e-05,
+      "loss": 1.8757,
+      "step": 11206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7829943064741165e-05,
+      "loss": 1.7419,
+      "step": 11207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7829570104881772e-05,
+      "loss": 1.7937,
+      "step": 11208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7829197116876927e-05,
+      "loss": 1.9592,
+      "step": 11209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.782882410072797e-05,
+      "loss": 1.7971,
+      "step": 11210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7828451056436243e-05,
+      "loss": 2.0957,
+      "step": 11211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7828077984003088e-05,
+      "loss": 2.0047,
+      "step": 11212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.782770488342984e-05,
+      "loss": 2.1964,
+      "step": 11213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7827331754717845e-05,
+      "loss": 1.8701,
+      "step": 11214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7826958597868444e-05,
+      "loss": 1.8398,
+      "step": 11215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7826585412882977e-05,
+      "loss": 1.9178,
+      "step": 11216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.782621219976279e-05,
+      "loss": 1.8369,
+      "step": 11217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7825838958509214e-05,
+      "loss": 2.1243,
+      "step": 11218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7825465689123605e-05,
+      "loss": 1.9793,
+      "step": 11219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7825092391607294e-05,
+      "loss": 2.1105,
+      "step": 11220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7824719065961624e-05,
+      "loss": 1.7945,
+      "step": 11221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7824345712187946e-05,
+      "loss": 1.8001,
+      "step": 11222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.782397233028759e-05,
+      "loss": 1.7021,
+      "step": 11223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7823598920261905e-05,
+      "loss": 2.1156,
+      "step": 11224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7823225482112235e-05,
+      "loss": 1.9069,
+      "step": 11225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7822852015839916e-05,
+      "loss": 1.8606,
+      "step": 11226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7822478521446294e-05,
+      "loss": 2.2397,
+      "step": 11227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.7822104998932715e-05,
+      "loss": 1.9397,
+      "step": 11228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7821731448300516e-05,
+      "loss": 1.9551,
+      "step": 11229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.782135786955104e-05,
+      "loss": 1.8577,
+      "step": 11230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7820984262685636e-05,
+      "loss": 1.7616,
+      "step": 11231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7820610627705642e-05,
+      "loss": 2.2448,
+      "step": 11232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.78202369646124e-05,
+      "loss": 1.5252,
+      "step": 11233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.781986327340726e-05,
+      "loss": 1.9348,
+      "step": 11234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7819489554091557e-05,
+      "loss": 1.6053,
+      "step": 11235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.781911580666664e-05,
+      "loss": 1.9916,
+      "step": 11236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7818742031133852e-05,
+      "loss": 2.0125,
+      "step": 11237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7818368227494535e-05,
+      "loss": 1.7662,
+      "step": 11238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.781799439575003e-05,
+      "loss": 1.9035,
+      "step": 11239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.781762053590169e-05,
+      "loss": 1.8228,
+      "step": 11240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.781724664795085e-05,
+      "loss": 1.8525,
+      "step": 11241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7816872731898854e-05,
+      "loss": 1.7632,
+      "step": 11242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7816498787747054e-05,
+      "loss": 1.6864,
+      "step": 11243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7816124815496785e-05,
+      "loss": 1.8878,
+      "step": 11244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.78157508151494e-05,
+      "loss": 1.9316,
+      "step": 11245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7815376786706234e-05,
+      "loss": 1.8138,
+      "step": 11246
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7815002730168642e-05,
+      "loss": 2.0604,
+      "step": 11247
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7814628645537962e-05,
+      "loss": 1.7595,
+      "step": 11248
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.781425453281554e-05,
+      "loss": 2.0275,
+      "step": 11249
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.781388039200272e-05,
+      "loss": 1.9747,
+      "step": 11250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7813506223100846e-05,
+      "loss": 2.0399,
+      "step": 11251
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7813132026111266e-05,
+      "loss": 2.1523,
+      "step": 11252
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7812757801035327e-05,
+      "loss": 1.6199,
+      "step": 11253
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.781238354787437e-05,
+      "loss": 1.7151,
+      "step": 11254
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7812009266629742e-05,
+      "loss": 2.025,
+      "step": 11255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7811634957302787e-05,
+      "loss": 1.7026,
+      "step": 11256
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7811260619894853e-05,
+      "loss": 1.7551,
+      "step": 11257
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7810886254407282e-05,
+      "loss": 1.8853,
+      "step": 11258
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7810511860841422e-05,
+      "loss": 1.7541,
+      "step": 11259
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7810137439198623e-05,
+      "loss": 2.1403,
+      "step": 11260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7809762989480222e-05,
+      "loss": 1.7448,
+      "step": 11261
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7809388511687575e-05,
+      "loss": 1.7582,
+      "step": 11262
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7809014005822023e-05,
+      "loss": 1.7268,
+      "step": 11263
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.780863947188491e-05,
+      "loss": 2.0097,
+      "step": 11264
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7808264909877583e-05,
+      "loss": 1.9808,
+      "step": 11265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7807890319801394e-05,
+      "loss": 1.7367,
+      "step": 11266
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7807515701657685e-05,
+      "loss": 1.9796,
+      "step": 11267
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.875,
+      "learning_rate": 1.7807141055447803e-05,
+      "loss": 1.9345,
+      "step": 11268
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7806766381173094e-05,
+      "loss": 1.8071,
+      "step": 11269
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7806391678834906e-05,
+      "loss": 1.7842,
+      "step": 11270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7806016948434588e-05,
+      "loss": 1.7856,
+      "step": 11271
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7805642189973484e-05,
+      "loss": 2.1586,
+      "step": 11272
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.780526740345294e-05,
+      "loss": 1.9356,
+      "step": 11273
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.780489258887431e-05,
+      "loss": 1.8641,
+      "step": 11274
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7804517746238935e-05,
+      "loss": 1.8409,
+      "step": 11275
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7804142875548165e-05,
+      "loss": 1.9503,
+      "step": 11276
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.780376797680335e-05,
+      "loss": 1.9657,
+      "step": 11277
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7803393050005827e-05,
+      "loss": 1.8093,
+      "step": 11278
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7803018095156957e-05,
+      "loss": 1.8039,
+      "step": 11279
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7802643112258082e-05,
+      "loss": 1.9573,
+      "step": 11280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7802268101310548e-05,
+      "loss": 2.0632,
+      "step": 11281
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7801893062315706e-05,
+      "loss": 2.1023,
+      "step": 11282
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7801517995274903e-05,
+      "loss": 1.8884,
+      "step": 11283
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.780114290018949e-05,
+      "loss": 2.0344,
+      "step": 11284
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7800767777060813e-05,
+      "loss": 2.0085,
+      "step": 11285
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.780039262589022e-05,
+      "loss": 1.9454,
+      "step": 11286
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.780001744667906e-05,
+      "loss": 1.8716,
+      "step": 11287
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7799642239428682e-05,
+      "loss": 1.9802,
+      "step": 11288
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7799267004140433e-05,
+      "loss": 2.3411,
+      "step": 11289
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7798891740815665e-05,
+      "loss": 1.8973,
+      "step": 11290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7798516449455726e-05,
+      "loss": 1.9158,
+      "step": 11291
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7798141130061967e-05,
+      "loss": 1.8036,
+      "step": 11292
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.779776578263573e-05,
+      "loss": 1.7506,
+      "step": 11293
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7797390407178373e-05,
+      "loss": 1.6813,
+      "step": 11294
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7797015003691238e-05,
+      "loss": 1.8486,
+      "step": 11295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7796639572175684e-05,
+      "loss": 2.0734,
+      "step": 11296
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.7796264112633047e-05,
+      "loss": 1.9587,
+      "step": 11297
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.779588862506469e-05,
+      "loss": 1.6374,
+      "step": 11298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7795513109471952e-05,
+      "loss": 2.3073,
+      "step": 11299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7795137565856193e-05,
+      "loss": 1.9824,
+      "step": 11300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7794761994218753e-05,
+      "loss": 1.556,
+      "step": 11301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7794386394560993e-05,
+      "loss": 1.7424,
+      "step": 11302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7794010766884254e-05,
+      "loss": 1.9831,
+      "step": 11303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7793635111189887e-05,
+      "loss": 1.9362,
+      "step": 11304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.779325942747925e-05,
+      "loss": 1.9305,
+      "step": 11305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7792883715753686e-05,
+      "loss": 2.1042,
+      "step": 11306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.779250797601455e-05,
+      "loss": 1.7369,
+      "step": 11307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7792132208263185e-05,
+      "loss": 2.0414,
+      "step": 11308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7791756412500953e-05,
+      "loss": 1.6543,
+      "step": 11309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7791380588729195e-05,
+      "loss": 2.3147,
+      "step": 11310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.779100473694927e-05,
+      "loss": 1.8379,
+      "step": 11311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7790628857162525e-05,
+      "loss": 2.0487,
+      "step": 11312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.7790252949370313e-05,
+      "loss": 1.9579,
+      "step": 11313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.778987701357398e-05,
+      "loss": 1.8557,
+      "step": 11314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7789501049774884e-05,
+      "loss": 1.8979,
+      "step": 11315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7789125057974374e-05,
+      "loss": 1.5414,
+      "step": 11316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.77887490381738e-05,
+      "loss": 2.1583,
+      "step": 11317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7788372990374518e-05,
+      "loss": 1.719,
+      "step": 11318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7787996914577874e-05,
+      "loss": 1.9643,
+      "step": 11319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7787620810785223e-05,
+      "loss": 2.062,
+      "step": 11320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.778724467899792e-05,
+      "loss": 1.7976,
+      "step": 11321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7786868519217312e-05,
+      "loss": 2.0567,
+      "step": 11322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7786492331444753e-05,
+      "loss": 2.0688,
+      "step": 11323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7786116115681596e-05,
+      "loss": 1.7609,
+      "step": 11324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7785739871929192e-05,
+      "loss": 1.8267,
+      "step": 11325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7785363600188894e-05,
+      "loss": 1.5754,
+      "step": 11326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.778498730046206e-05,
+      "loss": 1.9887,
+      "step": 11327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7784610972750028e-05,
+      "loss": 1.7359,
+      "step": 11328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.778423461705417e-05,
+      "loss": 1.9156,
+      "step": 11329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7783858233375825e-05,
+      "loss": 1.8088,
+      "step": 11330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.778348182171635e-05,
+      "loss": 2.0847,
+      "step": 11331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.77831053820771e-05,
+      "loss": 1.748,
+      "step": 11332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7782728914459428e-05,
+      "loss": 1.8128,
+      "step": 11333
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7782352418864684e-05,
+      "loss": 1.9649,
+      "step": 11334
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7781975895294223e-05,
+      "loss": 1.9794,
+      "step": 11335
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7781599343749402e-05,
+      "loss": 2.0553,
+      "step": 11336
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7781222764231572e-05,
+      "loss": 1.9351,
+      "step": 11337
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.778084615674208e-05,
+      "loss": 2.1634,
+      "step": 11338
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7780469521282293e-05,
+      "loss": 1.8597,
+      "step": 11339
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7780092857853553e-05,
+      "loss": 1.9981,
+      "step": 11340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7779716166457226e-05,
+      "loss": 2.5032,
+      "step": 11341
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7779339447094653e-05,
+      "loss": 1.9206,
+      "step": 11342
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7778962699767196e-05,
+      "loss": 1.7823,
+      "step": 11343
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.777858592447621e-05,
+      "loss": 1.9936,
+      "step": 11344
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7778209121223046e-05,
+      "loss": 1.96,
+      "step": 11345
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7777832290009057e-05,
+      "loss": 1.9855,
+      "step": 11346
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7777455430835605e-05,
+      "loss": 1.9102,
+      "step": 11347
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7777078543704035e-05,
+      "loss": 2.1111,
+      "step": 11348
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.777670162861571e-05,
+      "loss": 1.8304,
+      "step": 11349
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.777632468557198e-05,
+      "loss": 1.8245,
+      "step": 11350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7775947714574205e-05,
+      "loss": 1.6737,
+      "step": 11351
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.7775570715623735e-05,
+      "loss": 1.9371,
+      "step": 11352
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.7775193688721927e-05,
+      "loss": 2.1346,
+      "step": 11353
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7774816633870135e-05,
+      "loss": 2.0237,
+      "step": 11354
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.777443955106972e-05,
+      "loss": 1.7758,
+      "step": 11355
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.777406244032203e-05,
+      "loss": 1.9347,
+      "step": 11356
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7773685301628424e-05,
+      "loss": 2.3257,
+      "step": 11357
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.777330813499026e-05,
+      "loss": 1.8291,
+      "step": 11358
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7772930940408886e-05,
+      "loss": 2.0132,
+      "step": 11359
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.777255371788567e-05,
+      "loss": 1.863,
+      "step": 11360
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.777217646742196e-05,
+      "loss": 1.9232,
+      "step": 11361
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7771799189019114e-05,
+      "loss": 2.1827,
+      "step": 11362
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7771421882678487e-05,
+      "loss": 1.7482,
+      "step": 11363
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7771044548401436e-05,
+      "loss": 1.8834,
+      "step": 11364
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7770667186189316e-05,
+      "loss": 1.7833,
+      "step": 11365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7770289796043486e-05,
+      "loss": 1.7323,
+      "step": 11366
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7769912377965305e-05,
+      "loss": 1.681,
+      "step": 11367
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.776953493195612e-05,
+      "loss": 1.8743,
+      "step": 11368
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7769157458017303e-05,
+      "loss": 1.8436,
+      "step": 11369
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7768779956150196e-05,
+      "loss": 1.9004,
+      "step": 11370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.7768402426356165e-05,
+      "loss": 1.6239,
+      "step": 11371
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7768024868636566e-05,
+      "loss": 1.7923,
+      "step": 11372
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.776764728299275e-05,
+      "loss": 1.7407,
+      "step": 11373
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7767269669426083e-05,
+      "loss": 1.9075,
+      "step": 11374
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7766892027937914e-05,
+      "loss": 1.6364,
+      "step": 11375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.776651435852961e-05,
+      "loss": 1.8916,
+      "step": 11376
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7766136661202522e-05,
+      "loss": 1.8127,
+      "step": 11377
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7765758935958007e-05,
+      "loss": 1.9908,
+      "step": 11378
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7765381182797427e-05,
+      "loss": 1.6755,
+      "step": 11379
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7765003401722138e-05,
+      "loss": 1.7114,
+      "step": 11380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.77646255927335e-05,
+      "loss": 2.1057,
+      "step": 11381
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7764247755832866e-05,
+      "loss": 2.2563,
+      "step": 11382
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.77638698910216e-05,
+      "loss": 2.2156,
+      "step": 11383
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7763491998301057e-05,
+      "loss": 2.0569,
+      "step": 11384
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7763114077672594e-05,
+      "loss": 1.6646,
+      "step": 11385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7762736129137573e-05,
+      "loss": 1.9845,
+      "step": 11386
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7762358152697353e-05,
+      "loss": 1.9235,
+      "step": 11387
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.776198014835329e-05,
+      "loss": 1.8961,
+      "step": 11388
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7761602116106748e-05,
+      "loss": 1.607,
+      "step": 11389
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7761224055959077e-05,
+      "loss": 1.7546,
+      "step": 11390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.776084596791164e-05,
+      "loss": 1.9343,
+      "step": 11391
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7760467851965803e-05,
+      "loss": 1.8598,
+      "step": 11392
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7760089708122915e-05,
+      "loss": 2.0284,
+      "step": 11393
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.775971153638434e-05,
+      "loss": 1.9529,
+      "step": 11394
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7759333336751438e-05,
+      "loss": 1.9364,
+      "step": 11395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7758955109225567e-05,
+      "loss": 1.708,
+      "step": 11396
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.775857685380809e-05,
+      "loss": 1.9765,
+      "step": 11397
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7758198570500363e-05,
+      "loss": 1.6106,
+      "step": 11398
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7757820259303744e-05,
+      "loss": 2.0157,
+      "step": 11399
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.77574419202196e-05,
+      "loss": 2.031,
+      "step": 11400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7757063553249282e-05,
+      "loss": 2.01,
+      "step": 11401
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.775668515839416e-05,
+      "loss": 1.7173,
+      "step": 11402
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7756306735655583e-05,
+      "loss": 1.8372,
+      "step": 11403
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7755928285034922e-05,
+      "loss": 2.0725,
+      "step": 11404
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7755549806533534e-05,
+      "loss": 1.8883,
+      "step": 11405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.7755171300152775e-05,
+      "loss": 2.0597,
+      "step": 11406
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7754792765894014e-05,
+      "loss": 2.0267,
+      "step": 11407
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7754414203758602e-05,
+      "loss": 2.0898,
+      "step": 11408
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7754035613747906e-05,
+      "loss": 1.8765,
+      "step": 11409
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.775365699586329e-05,
+      "loss": 1.9198,
+      "step": 11410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7753278350106108e-05,
+      "loss": 1.7554,
+      "step": 11411
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.775289967647772e-05,
+      "loss": 1.4755,
+      "step": 11412
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7752520974979498e-05,
+      "loss": 1.8556,
+      "step": 11413
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7752142245612792e-05,
+      "loss": 1.7896,
+      "step": 11414
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.775176348837897e-05,
+      "loss": 2.1992,
+      "step": 11415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.775138470327939e-05,
+      "loss": 2.0072,
+      "step": 11416
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7751005890315416e-05,
+      "loss": 1.9147,
+      "step": 11417
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7750627049488408e-05,
+      "loss": 1.9616,
+      "step": 11418
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.775024818079973e-05,
+      "loss": 2.1773,
+      "step": 11419
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.774986928425074e-05,
+      "loss": 1.9813,
+      "step": 11420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.7749490359842807e-05,
+      "loss": 4.2628,
+      "step": 11421
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7749111407577285e-05,
+      "loss": 1.8094,
+      "step": 11422
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7748732427455545e-05,
+      "loss": 1.8779,
+      "step": 11423
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7748353419478938e-05,
+      "loss": 1.7581,
+      "step": 11424
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7747974383648837e-05,
+      "loss": 1.7903,
+      "step": 11425
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.77475953199666e-05,
+      "loss": 1.6943,
+      "step": 11426
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7747216228433587e-05,
+      "loss": 1.9189,
+      "step": 11427
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7746837109051167e-05,
+      "loss": 2.3512,
+      "step": 11428
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7746457961820703e-05,
+      "loss": 2.1772,
+      "step": 11429
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7746078786743547e-05,
+      "loss": 1.8684,
+      "step": 11430
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7745699583821073e-05,
+      "loss": 1.7339,
+      "step": 11431
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.774532035305464e-05,
+      "loss": 1.8335,
+      "step": 11432
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.774494109444562e-05,
+      "loss": 1.9989,
+      "step": 11433
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7744561807995356e-05,
+      "loss": 2.2601,
+      "step": 11434
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.774418249370523e-05,
+      "loss": 1.7245,
+      "step": 11435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.77438031515766e-05,
+      "loss": 1.8245,
+      "step": 11436
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.774342378161083e-05,
+      "loss": 1.61,
+      "step": 11437
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7743044383809276e-05,
+      "loss": 1.9278,
+      "step": 11438
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7742664958173316e-05,
+      "loss": 1.8106,
+      "step": 11439
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7742285504704304e-05,
+      "loss": 1.4782,
+      "step": 11440
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.774190602340361e-05,
+      "loss": 1.8696,
+      "step": 11441
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7741526514272585e-05,
+      "loss": 2.0045,
+      "step": 11442
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7741146977312613e-05,
+      "loss": 1.7224,
+      "step": 11443
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7740767412525043e-05,
+      "loss": 1.9579,
+      "step": 11444
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7740387819911245e-05,
+      "loss": 1.9134,
+      "step": 11445
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7740008199472588e-05,
+      "loss": 1.7673,
+      "step": 11446
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7739628551210426e-05,
+      "loss": 1.6778,
+      "step": 11447
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7739248875126132e-05,
+      "loss": 2.1639,
+      "step": 11448
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.773886917122107e-05,
+      "loss": 1.8136,
+      "step": 11449
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7738489439496604e-05,
+      "loss": 1.6596,
+      "step": 11450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7738109679954093e-05,
+      "loss": 1.9211,
+      "step": 11451
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7737729892594914e-05,
+      "loss": 2.0716,
+      "step": 11452
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7737350077420425e-05,
+      "loss": 1.8438,
+      "step": 11453
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.773697023443199e-05,
+      "loss": 1.7831,
+      "step": 11454
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.7736590363630978e-05,
+      "loss": 1.9126,
+      "step": 11455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7736210465018752e-05,
+      "loss": 1.9567,
+      "step": 11456
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.773583053859668e-05,
+      "loss": 1.9129,
+      "step": 11457
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7735450584366126e-05,
+      "loss": 1.8727,
+      "step": 11458
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.773507060232846e-05,
+      "loss": 1.7609,
+      "step": 11459
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.773469059248504e-05,
+      "loss": 1.9753,
+      "step": 11460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.773431055483724e-05,
+      "loss": 2.0691,
+      "step": 11461
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.773393048938642e-05,
+      "loss": 1.7934,
+      "step": 11462
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.773355039613395e-05,
+      "loss": 1.8501,
+      "step": 11463
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7733170275081193e-05,
+      "loss": 1.8986,
+      "step": 11464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.773279012622952e-05,
+      "loss": 1.9981,
+      "step": 11465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7732409949580297e-05,
+      "loss": 1.7216,
+      "step": 11466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7732029745134885e-05,
+      "loss": 1.7799,
+      "step": 11467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7731649512894657e-05,
+      "loss": 1.9758,
+      "step": 11468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7731269252860977e-05,
+      "loss": 1.9047,
+      "step": 11469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7730888965035215e-05,
+      "loss": 2.1483,
+      "step": 11470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.773050864941873e-05,
+      "loss": 1.7836,
+      "step": 11471
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7730128306012897e-05,
+      "loss": 2.3348,
+      "step": 11472
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7729747934819082e-05,
+      "loss": 1.8859,
+      "step": 11473
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.772936753583865e-05,
+      "loss": 1.9348,
+      "step": 11474
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.772898710907297e-05,
+      "loss": 1.7197,
+      "step": 11475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.772860665452341e-05,
+      "loss": 1.9179,
+      "step": 11476
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7728226172191338e-05,
+      "loss": 1.9621,
+      "step": 11477
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.772784566207812e-05,
+      "loss": 1.7557,
+      "step": 11478
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.772746512418512e-05,
+      "loss": 1.6211,
+      "step": 11479
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7727084558513713e-05,
+      "loss": 1.704,
+      "step": 11480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7726703965065266e-05,
+      "loss": 1.936,
+      "step": 11481
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7726323343841146e-05,
+      "loss": 2.0277,
+      "step": 11482
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.772594269484272e-05,
+      "loss": 1.7914,
+      "step": 11483
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7725562018071354e-05,
+      "loss": 2.1015,
+      "step": 11484
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7725181313528424e-05,
+      "loss": 1.8891,
+      "step": 11485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7724800581215292e-05,
+      "loss": 2.0467,
+      "step": 11486
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7724419821133327e-05,
+      "loss": 1.7959,
+      "step": 11487
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7724039033283904e-05,
+      "loss": 1.985,
+      "step": 11488
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.7723658217668385e-05,
+      "loss": 1.7294,
+      "step": 11489
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7723277374288142e-05,
+      "loss": 1.8816,
+      "step": 11490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7722896503144542e-05,
+      "loss": 1.6986,
+      "step": 11491
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7722515604238955e-05,
+      "loss": 1.5699,
+      "step": 11492
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.772213467757275e-05,
+      "loss": 1.629,
+      "step": 11493
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7721753723147303e-05,
+      "loss": 1.7961,
+      "step": 11494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.772137274096397e-05,
+      "loss": 1.99,
+      "step": 11495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7720991731024132e-05,
+      "loss": 2.081,
+      "step": 11496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7720610693329154e-05,
+      "loss": 1.8953,
+      "step": 11497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7720229627880407e-05,
+      "loss": 2.015,
+      "step": 11498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.771984853467926e-05,
+      "loss": 1.8268,
+      "step": 11499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7719467413727085e-05,
+      "loss": 1.9306,
+      "step": 11500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7719086265025248e-05,
+      "loss": 2.0615,
+      "step": 11501
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.771870508857512e-05,
+      "loss": 2.0293,
+      "step": 11502
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.7718323884378077e-05,
+      "loss": 2.048,
+      "step": 11503
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.771794265243548e-05,
+      "loss": 1.8799,
+      "step": 11504
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.7717561392748708e-05,
+      "loss": 2.2993,
+      "step": 11505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7717180105319124e-05,
+      "loss": 1.8303,
+      "step": 11506
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7716798790148106e-05,
+      "loss": 2.1856,
+      "step": 11507
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7716417447237023e-05,
+      "loss": 2.0866,
+      "step": 11508
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7716036076587238e-05,
+      "loss": 1.5108,
+      "step": 11509
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7715654678200132e-05,
+      "loss": 2.14,
+      "step": 11510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7715273252077075e-05,
+      "loss": 2.4653,
+      "step": 11511
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7714891798219432e-05,
+      "loss": 1.9464,
+      "step": 11512
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7714510316628578e-05,
+      "loss": 2.2348,
+      "step": 11513
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7714128807305883e-05,
+      "loss": 1.8207,
+      "step": 11514
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7713747270252717e-05,
+      "loss": 1.8261,
+      "step": 11515
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7713365705470458e-05,
+      "loss": 2.004,
+      "step": 11516
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7712984112960474e-05,
+      "loss": 1.7237,
+      "step": 11517
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7712602492724133e-05,
+      "loss": 2.006,
+      "step": 11518
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7712220844762812e-05,
+      "loss": 1.8333,
+      "step": 11519
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.771183916907788e-05,
+      "loss": 2.2802,
+      "step": 11520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.771145746567071e-05,
+      "loss": 1.5474,
+      "step": 11521
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.7711075734542675e-05,
+      "loss": 1.8898,
+      "step": 11522
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7710693975695145e-05,
+      "loss": 2.013,
+      "step": 11523
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.7710312189129496e-05,
+      "loss": 2.35,
+      "step": 11524
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7709930374847096e-05,
+      "loss": 1.7989,
+      "step": 11525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.770954853284932e-05,
+      "loss": 1.7801,
+      "step": 11526
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7709166663137537e-05,
+      "loss": 1.7087,
+      "step": 11527
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.770878476571313e-05,
+      "loss": 1.7998,
+      "step": 11528
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.770840284057746e-05,
+      "loss": 1.5714,
+      "step": 11529
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7708020887731907e-05,
+      "loss": 1.7829,
+      "step": 11530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.770763890717784e-05,
+      "loss": 1.7485,
+      "step": 11531
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.7707256898916634e-05,
+      "loss": 2.426,
+      "step": 11532
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7706874862949662e-05,
+      "loss": 1.6577,
+      "step": 11533
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7706492799278302e-05,
+      "loss": 2.5314,
+      "step": 11534
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7706110707903917e-05,
+      "loss": 1.9339,
+      "step": 11535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7705728588827888e-05,
+      "loss": 1.7491,
+      "step": 11536
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7705346442051588e-05,
+      "loss": 2.3358,
+      "step": 11537
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7704964267576392e-05,
+      "loss": 1.6021,
+      "step": 11538
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.7704582065403667e-05,
+      "loss": 1.6205,
+      "step": 11539
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7704199835534793e-05,
+      "loss": 2.0381,
+      "step": 11540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7703817577971144e-05,
+      "loss": 1.8952,
+      "step": 11541
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.770343529271409e-05,
+      "loss": 1.9556,
+      "step": 11542
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7703052979765014e-05,
+      "loss": 1.7845,
+      "step": 11543
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.770267063912528e-05,
+      "loss": 1.6922,
+      "step": 11544
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7702288270796266e-05,
+      "loss": 1.655,
+      "step": 11545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.770190587477935e-05,
+      "loss": 1.9486,
+      "step": 11546
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7701523451075903e-05,
+      "loss": 2.0097,
+      "step": 11547
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7701140999687297e-05,
+      "loss": 1.9205,
+      "step": 11548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7700758520614913e-05,
+      "loss": 1.6015,
+      "step": 11549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7700376013860124e-05,
+      "loss": 2.0123,
+      "step": 11550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7699993479424305e-05,
+      "loss": 2.1825,
+      "step": 11551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7699610917308828e-05,
+      "loss": 1.9796,
+      "step": 11552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7699228327515072e-05,
+      "loss": 1.7028,
+      "step": 11553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.769884571004441e-05,
+      "loss": 1.7032,
+      "step": 11554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.769846306489822e-05,
+      "loss": 1.8042,
+      "step": 11555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.769808039207787e-05,
+      "loss": 1.7938,
+      "step": 11556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.769769769158475e-05,
+      "loss": 1.9308,
+      "step": 11557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7697314963420223e-05,
+      "loss": 2.0515,
+      "step": 11558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7696932207585672e-05,
+      "loss": 2.0596,
+      "step": 11559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7696549424082468e-05,
+      "loss": 1.9465,
+      "step": 11560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.769616661291199e-05,
+      "loss": 1.941,
+      "step": 11561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.75,
+      "learning_rate": 1.769578377407561e-05,
+      "loss": 2.2579,
+      "step": 11562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.769540090757471e-05,
+      "loss": 1.9006,
+      "step": 11563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.769501801341066e-05,
+      "loss": 2.0995,
+      "step": 11564
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7694635091584845e-05,
+      "loss": 2.3383,
+      "step": 11565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7694252142098633e-05,
+      "loss": 2.0063,
+      "step": 11566
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7693869164953407e-05,
+      "loss": 2.0291,
+      "step": 11567
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7693486160150543e-05,
+      "loss": 1.715,
+      "step": 11568
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.769310312769141e-05,
+      "loss": 1.7353,
+      "step": 11569
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7692720067577392e-05,
+      "loss": 2.0035,
+      "step": 11570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7692336979809866e-05,
+      "loss": 1.7957,
+      "step": 11571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7691953864390208e-05,
+      "loss": 1.9105,
+      "step": 11572
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.7691570721319797e-05,
+      "loss": 1.5294,
+      "step": 11573
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7691187550600003e-05,
+      "loss": 2.0895,
+      "step": 11574
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.769080435223221e-05,
+      "loss": 1.7646,
+      "step": 11575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7690421126217796e-05,
+      "loss": 1.8395,
+      "step": 11576
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7690037872558136e-05,
+      "loss": 2.0473,
+      "step": 11577
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.768965459125461e-05,
+      "loss": 1.583,
+      "step": 11578
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.768927128230859e-05,
+      "loss": 1.5715,
+      "step": 11579
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7688887945721464e-05,
+      "loss": 1.8515,
+      "step": 11580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7688504581494603e-05,
+      "loss": 1.9509,
+      "step": 11581
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7688121189629384e-05,
+      "loss": 1.6555,
+      "step": 11582
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.768773777012719e-05,
+      "loss": 1.9269,
+      "step": 11583
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.768735432298939e-05,
+      "loss": 1.6953,
+      "step": 11584
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7686970848217377e-05,
+      "loss": 1.9886,
+      "step": 11585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.768658734581252e-05,
+      "loss": 1.6597,
+      "step": 11586
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7686203815776197e-05,
+      "loss": 2.0625,
+      "step": 11587
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7685820258109788e-05,
+      "loss": 2.1678,
+      "step": 11588
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.768543667281468e-05,
+      "loss": 1.7379,
+      "step": 11589
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7685053059892235e-05,
+      "loss": 2.141,
+      "step": 11590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.768466941934385e-05,
+      "loss": 1.8202,
+      "step": 11591
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.768428575117089e-05,
+      "loss": 1.772,
+      "step": 11592
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7683902055374746e-05,
+      "loss": 1.9159,
+      "step": 11593
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7683518331956783e-05,
+      "loss": 2.0394,
+      "step": 11594
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7683134580918394e-05,
+      "loss": 1.8289,
+      "step": 11595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7682750802260953e-05,
+      "loss": 2.1941,
+      "step": 11596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.768236699598584e-05,
+      "loss": 1.9453,
+      "step": 11597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.7681983162094433e-05,
+      "loss": 1.9781,
+      "step": 11598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7681599300588114e-05,
+      "loss": 1.9711,
+      "step": 11599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7681215411468264e-05,
+      "loss": 1.7742,
+      "step": 11600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.768083149473626e-05,
+      "loss": 1.9322,
+      "step": 11601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7680447550393484e-05,
+      "loss": 2.0248,
+      "step": 11602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.768006357844131e-05,
+      "loss": 2.1728,
+      "step": 11603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.767967957888113e-05,
+      "loss": 1.7652,
+      "step": 11604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7679295551714316e-05,
+      "loss": 1.9525,
+      "step": 11605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7678911496942248e-05,
+      "loss": 1.9118,
+      "step": 11606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7678527414566314e-05,
+      "loss": 2.1619,
+      "step": 11607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7678143304587888e-05,
+      "loss": 1.9945,
+      "step": 11608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.767775916700835e-05,
+      "loss": 1.6865,
+      "step": 11609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.7677375001829086e-05,
+      "loss": 2.0516,
+      "step": 11610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7676990809051475e-05,
+      "loss": 2.0177,
+      "step": 11611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7676606588676896e-05,
+      "loss": 2.2106,
+      "step": 11612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7676222340706733e-05,
+      "loss": 1.8009,
+      "step": 11613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7675838065142365e-05,
+      "loss": 1.9156,
+      "step": 11614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7675453761985176e-05,
+      "loss": 1.5748,
+      "step": 11615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7675069431236543e-05,
+      "loss": 2.0024,
+      "step": 11616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7674685072897855e-05,
+      "loss": 1.7304,
+      "step": 11617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7674300686970488e-05,
+      "loss": 1.8991,
+      "step": 11618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7673916273455823e-05,
+      "loss": 2.0532,
+      "step": 11619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7673531832355243e-05,
+      "loss": 1.9883,
+      "step": 11620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7673147363670137e-05,
+      "loss": 1.7531,
+      "step": 11621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.7672762867401874e-05,
+      "loss": 2.0397,
+      "step": 11622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7672378343551844e-05,
+      "loss": 1.7816,
+      "step": 11623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7671993792121428e-05,
+      "loss": 1.8381,
+      "step": 11624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7671609213112012e-05,
+      "loss": 1.8563,
+      "step": 11625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7671224606524972e-05,
+      "loss": 1.8935,
+      "step": 11626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7670839972361697e-05,
+      "loss": 1.8239,
+      "step": 11627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7670455310623563e-05,
+      "loss": 1.755,
+      "step": 11628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.767007062131196e-05,
+      "loss": 1.8338,
+      "step": 11629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7669685904428263e-05,
+      "loss": 2.3162,
+      "step": 11630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.766930115997386e-05,
+      "loss": 1.7596,
+      "step": 11631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7668916387950136e-05,
+      "loss": 1.8966,
+      "step": 11632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7668531588358467e-05,
+      "loss": 1.5041,
+      "step": 11633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7668146761200243e-05,
+      "loss": 1.9435,
+      "step": 11634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7667761906476846e-05,
+      "loss": 1.8611,
+      "step": 11635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7667377024189655e-05,
+      "loss": 1.6504,
+      "step": 11636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.766699211434006e-05,
+      "loss": 2.0039,
+      "step": 11637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.766660717692944e-05,
+      "loss": 1.8951,
+      "step": 11638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.766622221195918e-05,
+      "loss": 1.8826,
+      "step": 11639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7665837219430665e-05,
+      "loss": 1.8615,
+      "step": 11640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7665452199345278e-05,
+      "loss": 1.7475,
+      "step": 11641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.76650671517044e-05,
+      "loss": 2.0492,
+      "step": 11642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7664682076509424e-05,
+      "loss": 1.9106,
+      "step": 11643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7664296973761726e-05,
+      "loss": 1.8715,
+      "step": 11644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7663911843462692e-05,
+      "loss": 1.88,
+      "step": 11645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7663526685613708e-05,
+      "loss": 2.2756,
+      "step": 11646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7663141500216157e-05,
+      "loss": 1.8247,
+      "step": 11647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7662756287271427e-05,
+      "loss": 1.8671,
+      "step": 11648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.76623710467809e-05,
+      "loss": 2.0375,
+      "step": 11649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.766198577874596e-05,
+      "loss": 2.0205,
+      "step": 11650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7661600483167993e-05,
+      "loss": 2.7319,
+      "step": 11651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7661215160048384e-05,
+      "loss": 1.916,
+      "step": 11652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.766082980938852e-05,
+      "loss": 1.6778,
+      "step": 11653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 1.8107,
+      "step": 11654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.766005902545356e-05,
+      "loss": 2.0449,
+      "step": 11655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7659673592181238e-05,
+      "loss": 2.0312,
+      "step": 11656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7659288131374196e-05,
+      "loss": 1.5399,
+      "step": 11657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.765890264303383e-05,
+      "loss": 2.2315,
+      "step": 11658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.765851712716152e-05,
+      "loss": 1.8566,
+      "step": 11659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.765813158375865e-05,
+      "loss": 2.0049,
+      "step": 11660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7657746012826608e-05,
+      "loss": 1.7956,
+      "step": 11661
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7657360414366778e-05,
+      "loss": 1.9552,
+      "step": 11662
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7656974788380553e-05,
+      "loss": 1.4478,
+      "step": 11663
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7656589134869312e-05,
+      "loss": 2.213,
+      "step": 11664
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7656203453834443e-05,
+      "loss": 1.7176,
+      "step": 11665
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7655817745277335e-05,
+      "loss": 1.9125,
+      "step": 11666
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7655432009199372e-05,
+      "loss": 1.9288,
+      "step": 11667
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7655046245601942e-05,
+      "loss": 2.0142,
+      "step": 11668
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.765466045448643e-05,
+      "loss": 1.7474,
+      "step": 11669
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7654274635854224e-05,
+      "loss": 1.8639,
+      "step": 11670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.7653888789706714e-05,
+      "loss": 2.1365,
+      "step": 11671
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.765350291604528e-05,
+      "loss": 1.9222,
+      "step": 11672
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7653117014871312e-05,
+      "loss": 2.1861,
+      "step": 11673
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.76527310861862e-05,
+      "loss": 1.8629,
+      "step": 11674
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.765234512999133e-05,
+      "loss": 2.0336,
+      "step": 11675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.765195914628809e-05,
+      "loss": 1.6032,
+      "step": 11676
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.7651573135077867e-05,
+      "loss": 2.1619,
+      "step": 11677
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.765118709636205e-05,
+      "loss": 1.8613,
+      "step": 11678
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.765080103014202e-05,
+      "loss": 2.1113,
+      "step": 11679
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.765041493641917e-05,
+      "loss": 2.0843,
+      "step": 11680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.765002881519489e-05,
+      "loss": 1.781,
+      "step": 11681
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.375,
+      "learning_rate": 1.764964266647056e-05,
+      "loss": 1.8932,
+      "step": 11682
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.764925649024758e-05,
+      "loss": 1.8732,
+      "step": 11683
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.764887028652733e-05,
+      "loss": 1.9436,
+      "step": 11684
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7648484055311202e-05,
+      "loss": 1.8645,
+      "step": 11685
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.764809779660058e-05,
+      "loss": 1.7801,
+      "step": 11686
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7647711510396854e-05,
+      "loss": 1.9309,
+      "step": 11687
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7647325196701418e-05,
+      "loss": 1.5885,
+      "step": 11688
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7646938855515653e-05,
+      "loss": 1.5168,
+      "step": 11689
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.764655248684095e-05,
+      "loss": 1.8854,
+      "step": 11690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7646166090678706e-05,
+      "loss": 1.7272,
+      "step": 11691
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.76457796670303e-05,
+      "loss": 1.8711,
+      "step": 11692
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.764539321589712e-05,
+      "loss": 1.7245,
+      "step": 11693
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.764500673728056e-05,
+      "loss": 1.8358,
+      "step": 11694
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7644620231182015e-05,
+      "loss": 2.0052,
+      "step": 11695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7644233697602864e-05,
+      "loss": 1.7666,
+      "step": 11696
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.76438471365445e-05,
+      "loss": 1.8774,
+      "step": 11697
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7643460548008316e-05,
+      "loss": 2.2834,
+      "step": 11698
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.76430739319957e-05,
+      "loss": 1.8447,
+      "step": 11699
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7642687288508035e-05,
+      "loss": 1.8028,
+      "step": 11700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7642300617546723e-05,
+      "loss": 2.0347,
+      "step": 11701
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.764191391911314e-05,
+      "loss": 1.8691,
+      "step": 11702
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.764152719320869e-05,
+      "loss": 2.3312,
+      "step": 11703
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7641140439834755e-05,
+      "loss": 1.9521,
+      "step": 11704
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7640753658992728e-05,
+      "loss": 1.915,
+      "step": 11705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7640366850684e-05,
+      "loss": 1.9511,
+      "step": 11706
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.7639980014909954e-05,
+      "loss": 2.0139,
+      "step": 11707
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7639593151671993e-05,
+      "loss": 1.734,
+      "step": 11708
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7639206260971497e-05,
+      "loss": 1.8256,
+      "step": 11709
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7638819342809866e-05,
+      "loss": 1.7388,
+      "step": 11710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.7638432397188482e-05,
+      "loss": 1.8311,
+      "step": 11711
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7638045424108742e-05,
+      "loss": 2.0458,
+      "step": 11712
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7637658423572033e-05,
+      "loss": 1.7692,
+      "step": 11713
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.763727139557975e-05,
+      "loss": 1.7453,
+      "step": 11714
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.763688434013328e-05,
+      "loss": 1.9347,
+      "step": 11715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7636497257234018e-05,
+      "loss": 1.824,
+      "step": 11716
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7636110146883356e-05,
+      "loss": 1.8507,
+      "step": 11717
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.763572300908268e-05,
+      "loss": 2.1907,
+      "step": 11718
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.763533584383339e-05,
+      "loss": 1.7911,
+      "step": 11719
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.763494865113687e-05,
+      "loss": 2.2893,
+      "step": 11720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.763456143099452e-05,
+      "loss": 1.9844,
+      "step": 11721
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7634174183407722e-05,
+      "loss": 1.8472,
+      "step": 11722
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7633786908377877e-05,
+      "loss": 1.3423,
+      "step": 11723
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.763339960590637e-05,
+      "loss": 1.8473,
+      "step": 11724
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.76330122759946e-05,
+      "loss": 1.9971,
+      "step": 11725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7632624918643955e-05,
+      "loss": 1.7261,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.763223753385583e-05,
+      "loss": 1.9742,
+      "step": 11727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.7631850121631613e-05,
+      "loss": 1.6112,
+      "step": 11728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.76314626819727e-05,
+      "loss": 1.9237,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7631075214880485e-05,
+      "loss": 1.8217,
+      "step": 11730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7630687720356356e-05,
+      "loss": 1.6215,
+      "step": 11731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7630300198401713e-05,
+      "loss": 1.8776,
+      "step": 11732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7629912649017944e-05,
+      "loss": 1.8264,
+      "step": 11733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7629525072206445e-05,
+      "loss": 1.6906,
+      "step": 11734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7629137467968605e-05,
+      "loss": 1.7989,
+      "step": 11735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.762874983630582e-05,
+      "loss": 1.7219,
+      "step": 11736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7628362177219487e-05,
+      "loss": 1.9,
+      "step": 11737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7627974490710994e-05,
+      "loss": 2.0989,
+      "step": 11738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.7627586776781734e-05,
+      "loss": 1.7845,
+      "step": 11739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7627199035433106e-05,
+      "loss": 1.4937,
+      "step": 11740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7626811266666503e-05,
+      "loss": 2.0038,
+      "step": 11741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7626423470483313e-05,
+      "loss": 1.7629,
+      "step": 11742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.762603564688494e-05,
+      "loss": 1.8177,
+      "step": 11743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7625647795872765e-05,
+      "loss": 1.5851,
+      "step": 11744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7625259917448195e-05,
+      "loss": 2.0562,
+      "step": 11745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7624872011612616e-05,
+      "loss": 1.7239,
+      "step": 11746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7624484078367427e-05,
+      "loss": 1.9788,
+      "step": 11747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7624096117714024e-05,
+      "loss": 1.7296,
+      "step": 11748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7623708129653792e-05,
+      "loss": 1.852,
+      "step": 11749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7623320114188135e-05,
+      "loss": 1.6788,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7622932071318447e-05,
+      "loss": 1.8034,
+      "step": 11751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.762254400104612e-05,
+      "loss": 1.4598,
+      "step": 11752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7622155903372547e-05,
+      "loss": 2.0428,
+      "step": 11753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7621767778299128e-05,
+      "loss": 1.8715,
+      "step": 11754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7621379625827257e-05,
+      "loss": 1.6096,
+      "step": 11755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.762099144595833e-05,
+      "loss": 1.9432,
+      "step": 11756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.762060323869374e-05,
+      "loss": 2.0307,
+      "step": 11757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.762021500403488e-05,
+      "loss": 1.6774,
+      "step": 11758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7619826741983153e-05,
+      "loss": 2.4789,
+      "step": 11759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.761943845253995e-05,
+      "loss": 1.7355,
+      "step": 11760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.761905013570667e-05,
+      "loss": 2.0907,
+      "step": 11761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7618661791484704e-05,
+      "loss": 1.7195,
+      "step": 11762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7618273419875448e-05,
+      "loss": 1.9536,
+      "step": 11763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7617885020880306e-05,
+      "loss": 1.9066,
+      "step": 11764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.7617496594500666e-05,
+      "loss": 1.8339,
+      "step": 11765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.761710814073793e-05,
+      "loss": 1.9938,
+      "step": 11766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.761671965959349e-05,
+      "loss": 2.0081,
+      "step": 11767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.761633115106874e-05,
+      "loss": 1.914,
+      "step": 11768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7615942615165088e-05,
+      "loss": 1.8029,
+      "step": 11769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7615554051883918e-05,
+      "loss": 1.4865,
+      "step": 11770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7615165461226636e-05,
+      "loss": 1.8067,
+      "step": 11771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.761477684319463e-05,
+      "loss": 1.7963,
+      "step": 11772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7614388197789307e-05,
+      "loss": 1.8648,
+      "step": 11773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.761399952501206e-05,
+      "loss": 2.3253,
+      "step": 11774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.761361082486428e-05,
+      "loss": 2.1645,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7613222097347372e-05,
+      "loss": 1.8449,
+      "step": 11776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.761283334246273e-05,
+      "loss": 1.7787,
+      "step": 11777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7612444560211757e-05,
+      "loss": 1.9762,
+      "step": 11778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7612055750595845e-05,
+      "loss": 1.7462,
+      "step": 11779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.761166691361639e-05,
+      "loss": 1.7751,
+      "step": 11780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.761127804927479e-05,
+      "loss": 2.2137,
+      "step": 11781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7610889157572453e-05,
+      "loss": 1.907,
+      "step": 11782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7610500238510763e-05,
+      "loss": 1.4477,
+      "step": 11783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.761011129209113e-05,
+      "loss": 1.6733,
+      "step": 11784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7609722318314943e-05,
+      "loss": 2.1318,
+      "step": 11785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.76093333171836e-05,
+      "loss": 1.9364,
+      "step": 11786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.760894428869851e-05,
+      "loss": 1.9683,
+      "step": 11787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7608555232861063e-05,
+      "loss": 1.8758,
+      "step": 11788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.760816614967266e-05,
+      "loss": 1.7515,
+      "step": 11789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7607777039134696e-05,
+      "loss": 1.8356,
+      "step": 11790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7607387901248573e-05,
+      "loss": 2.2906,
+      "step": 11791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7606998736015694e-05,
+      "loss": 1.8595,
+      "step": 11792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7606609543437446e-05,
+      "loss": 2.4181,
+      "step": 11793
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7606220323515244e-05,
+      "loss": 1.7605,
+      "step": 11794
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7605831076250473e-05,
+      "loss": 1.9654,
+      "step": 11795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.760544180164454e-05,
+      "loss": 1.8554,
+      "step": 11796
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.760505249969884e-05,
+      "loss": 1.7559,
+      "step": 11797
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.375,
+      "learning_rate": 1.760466317041478e-05,
+      "loss": 1.7589,
+      "step": 11798
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.760427381379375e-05,
+      "loss": 2.0942,
+      "step": 11799
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7603884429837154e-05,
+      "loss": 1.9445,
+      "step": 11800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7603495018546394e-05,
+      "loss": 1.8025,
+      "step": 11801
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7603105579922867e-05,
+      "loss": 1.7795,
+      "step": 11802
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7602716113967975e-05,
+      "loss": 1.6345,
+      "step": 11803
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7602326620683115e-05,
+      "loss": 1.6797,
+      "step": 11804
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7601937100069688e-05,
+      "loss": 1.9202,
+      "step": 11805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.76015475521291e-05,
+      "loss": 1.8421,
+      "step": 11806
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7601157976862738e-05,
+      "loss": 1.8472,
+      "step": 11807
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7600768374272013e-05,
+      "loss": 1.6256,
+      "step": 11808
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7600378744358325e-05,
+      "loss": 2.2075,
+      "step": 11809
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7599989087123072e-05,
+      "loss": 1.8021,
+      "step": 11810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.759959940256766e-05,
+      "loss": 2.1762,
+      "step": 11811
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7599209690693478e-05,
+      "loss": 1.599,
+      "step": 11812
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.759881995150194e-05,
+      "loss": 2.0418,
+      "step": 11813
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7598430184994438e-05,
+      "loss": 1.7881,
+      "step": 11814
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7598040391172374e-05,
+      "loss": 1.6378,
+      "step": 11815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.759765057003716e-05,
+      "loss": 2.0868,
+      "step": 11816
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7597260721590182e-05,
+      "loss": 1.5731,
+      "step": 11817
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.759687084583285e-05,
+      "loss": 2.0639,
+      "step": 11818
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7596480942766568e-05,
+      "loss": 2.1329,
+      "step": 11819
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.759609101239273e-05,
+      "loss": 1.8973,
+      "step": 11820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.759570105471274e-05,
+      "loss": 1.9396,
+      "step": 11821
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7595311069728e-05,
+      "loss": 1.9669,
+      "step": 11822
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7594921057439917e-05,
+      "loss": 2.0635,
+      "step": 11823
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.759453101784989e-05,
+      "loss": 1.7125,
+      "step": 11824
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7594140950959313e-05,
+      "loss": 2.0451,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7593750856769602e-05,
+      "loss": 1.7391,
+      "step": 11826
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.759336073528215e-05,
+      "loss": 1.7124,
+      "step": 11827
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7592970586498363e-05,
+      "loss": 1.6481,
+      "step": 11828
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.759258041041964e-05,
+      "loss": 1.9232,
+      "step": 11829
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.759219020704739e-05,
+      "loss": 1.9739,
+      "step": 11830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7591799976383007e-05,
+      "loss": 1.6931,
+      "step": 11831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.75914097184279e-05,
+      "loss": 1.7818,
+      "step": 11832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.759101943318347e-05,
+      "loss": 1.9588,
+      "step": 11833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7590629120651122e-05,
+      "loss": 1.79,
+      "step": 11834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7590238780832258e-05,
+      "loss": 1.8998,
+      "step": 11835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.758984841372828e-05,
+      "loss": 1.9116,
+      "step": 11836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.758945801934059e-05,
+      "loss": 1.6615,
+      "step": 11837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.75890675976706e-05,
+      "loss": 1.9754,
+      "step": 11838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7588677148719697e-05,
+      "loss": 2.0089,
+      "step": 11839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.75882866724893e-05,
+      "loss": 2.0113,
+      "step": 11840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7587896168980803e-05,
+      "loss": 1.9673,
+      "step": 11841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.758750563819562e-05,
+      "loss": 1.6895,
+      "step": 11842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7587115080135145e-05,
+      "loss": 2.1097,
+      "step": 11843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.7586724494800786e-05,
+      "loss": 2.1488,
+      "step": 11844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7586333882193944e-05,
+      "loss": 1.6079,
+      "step": 11845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.758594324231603e-05,
+      "loss": 1.3097,
+      "step": 11846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.758555257516844e-05,
+      "loss": 1.6387,
+      "step": 11847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7585161880752586e-05,
+      "loss": 1.9268,
+      "step": 11848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7584771159069866e-05,
+      "loss": 1.8676,
+      "step": 11849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.758438041012169e-05,
+      "loss": 1.7291,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.758398963390946e-05,
+      "loss": 1.9337,
+      "step": 11851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.375,
+      "learning_rate": 1.758359883043458e-05,
+      "loss": 1.8856,
+      "step": 11852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7583207999698456e-05,
+      "loss": 2.061,
+      "step": 11853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.758281714170249e-05,
+      "loss": 1.7057,
+      "step": 11854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7582426256448092e-05,
+      "loss": 1.7056,
+      "step": 11855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7582035343936666e-05,
+      "loss": 1.9477,
+      "step": 11856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7581644404169612e-05,
+      "loss": 1.7802,
+      "step": 11857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7581253437148346e-05,
+      "loss": 1.7826,
+      "step": 11858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.758086244287426e-05,
+      "loss": 1.941,
+      "step": 11859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.758047142134877e-05,
+      "loss": 1.7589,
+      "step": 11860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.758008037257328e-05,
+      "loss": 1.8639,
+      "step": 11861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.757968929654919e-05,
+      "loss": 1.9648,
+      "step": 11862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.757929819327791e-05,
+      "loss": 1.7215,
+      "step": 11863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7578907062760846e-05,
+      "loss": 2.0816,
+      "step": 11864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7578515904999402e-05,
+      "loss": 2.0051,
+      "step": 11865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7578124719994986e-05,
+      "loss": 1.8595,
+      "step": 11866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7577733507749007e-05,
+      "loss": 1.846,
+      "step": 11867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7577342268262866e-05,
+      "loss": 2.0325,
+      "step": 11868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.757695100153797e-05,
+      "loss": 2.1503,
+      "step": 11869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.757655970757573e-05,
+      "loss": 1.9315,
+      "step": 11870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7576168386377547e-05,
+      "loss": 1.762,
+      "step": 11871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7575777037944832e-05,
+      "loss": 2.1937,
+      "step": 11872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7575385662278988e-05,
+      "loss": 2.0675,
+      "step": 11873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7574994259381428e-05,
+      "loss": 1.7507,
+      "step": 11874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.757460282925355e-05,
+      "loss": 1.6952,
+      "step": 11875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.757421137189677e-05,
+      "loss": 1.9499,
+      "step": 11876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.757381988731249e-05,
+      "loss": 1.946,
+      "step": 11877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.757342837550212e-05,
+      "loss": 1.9801,
+      "step": 11878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7573036836467063e-05,
+      "loss": 1.788,
+      "step": 11879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7572645270208736e-05,
+      "loss": 1.8782,
+      "step": 11880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7572253676728533e-05,
+      "loss": 1.9868,
+      "step": 11881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.757186205602787e-05,
+      "loss": 2.1207,
+      "step": 11882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7571470408108156e-05,
+      "loss": 1.5952,
+      "step": 11883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7571078732970798e-05,
+      "loss": 1.9384,
+      "step": 11884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.7570687030617197e-05,
+      "loss": 2.1469,
+      "step": 11885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.757029530104877e-05,
+      "loss": 1.8251,
+      "step": 11886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7569903544266923e-05,
+      "loss": 1.9446,
+      "step": 11887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7569511760273062e-05,
+      "loss": 1.8387,
+      "step": 11888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.7569119949068592e-05,
+      "loss": 2.3208,
+      "step": 11889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.756872811065493e-05,
+      "loss": 2.1292,
+      "step": 11890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.756833624503348e-05,
+      "loss": 1.8997,
+      "step": 11891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7567944352205653e-05,
+      "loss": 1.8799,
+      "step": 11892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7567552432172852e-05,
+      "loss": 2.0506,
+      "step": 11893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7567160484936494e-05,
+      "loss": 1.8955,
+      "step": 11894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.756676851049798e-05,
+      "loss": 1.7549,
+      "step": 11895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.7566376508858725e-05,
+      "loss": 2.2767,
+      "step": 11896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7565984480020134e-05,
+      "loss": 1.7221,
+      "step": 11897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7565592423983617e-05,
+      "loss": 1.9854,
+      "step": 11898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7565200340750586e-05,
+      "loss": 1.9465,
+      "step": 11899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.756480823032245e-05,
+      "loss": 1.7234,
+      "step": 11900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.756441609270062e-05,
+      "loss": 1.8188,
+      "step": 11901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7564023927886497e-05,
+      "loss": 1.7451,
+      "step": 11902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.75636317358815e-05,
+      "loss": 2.0979,
+      "step": 11903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7563239516687033e-05,
+      "loss": 1.9704,
+      "step": 11904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.756284727030451e-05,
+      "loss": 2.2043,
+      "step": 11905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7562454996735342e-05,
+      "loss": 1.7284,
+      "step": 11906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7562062695980933e-05,
+      "loss": 1.8428,
+      "step": 11907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.75616703680427e-05,
+      "loss": 1.9209,
+      "step": 11908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.756127801292205e-05,
+      "loss": 2.1164,
+      "step": 11909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7560885630620395e-05,
+      "loss": 1.7609,
+      "step": 11910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.756049322113914e-05,
+      "loss": 2.1391,
+      "step": 11911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7560100784479705e-05,
+      "loss": 1.6372,
+      "step": 11912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.755970832064349e-05,
+      "loss": 1.8407,
+      "step": 11913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7559315829631917e-05,
+      "loss": 2.0275,
+      "step": 11914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7558923311446386e-05,
+      "loss": 1.7963,
+      "step": 11915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7558530766088317e-05,
+      "loss": 1.8886,
+      "step": 11916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7558138193559115e-05,
+      "loss": 2.0541,
+      "step": 11917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7557745593860196e-05,
+      "loss": 2.0815,
+      "step": 11918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.755735296699297e-05,
+      "loss": 1.7143,
+      "step": 11919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7556960312958842e-05,
+      "loss": 2.1914,
+      "step": 11920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7556567631759233e-05,
+      "loss": 1.862,
+      "step": 11921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.755617492339555e-05,
+      "loss": 1.902,
+      "step": 11922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7555782187869207e-05,
+      "loss": 1.9127,
+      "step": 11923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.755538942518161e-05,
+      "loss": 1.9425,
+      "step": 11924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7554996635334175e-05,
+      "loss": 1.7005,
+      "step": 11925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7554603818328316e-05,
+      "loss": 2.172,
+      "step": 11926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7554210974165443e-05,
+      "loss": 1.9249,
+      "step": 11927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7553818102846968e-05,
+      "loss": 1.9633,
+      "step": 11928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.75534252043743e-05,
+      "loss": 2.3231,
+      "step": 11929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.7553032278748857e-05,
+      "loss": 1.3919,
+      "step": 11930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7552639325972053e-05,
+      "loss": 1.9717,
+      "step": 11931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7552246346045292e-05,
+      "loss": 1.8667,
+      "step": 11932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.755185333896999e-05,
+      "loss": 1.8382,
+      "step": 11933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7551460304747564e-05,
+      "loss": 1.8783,
+      "step": 11934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7551067243379425e-05,
+      "loss": 2.0812,
+      "step": 11935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7550674154866984e-05,
+      "loss": 1.5661,
+      "step": 11936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7550281039211654e-05,
+      "loss": 1.8408,
+      "step": 11937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7549887896414853e-05,
+      "loss": 2.299,
+      "step": 11938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7549494726477986e-05,
+      "loss": 2.1507,
+      "step": 11939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7549101529402474e-05,
+      "loss": 1.7793,
+      "step": 11940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7548708305189724e-05,
+      "loss": 1.7346,
+      "step": 11941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7548315053841155e-05,
+      "loss": 1.7721,
+      "step": 11942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7547921775358177e-05,
+      "loss": 1.6381,
+      "step": 11943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.754752846974221e-05,
+      "loss": 1.6144,
+      "step": 11944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.754713513699466e-05,
+      "loss": 1.6529,
+      "step": 11945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.754674177711694e-05,
+      "loss": 1.6352,
+      "step": 11946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7546348390110476e-05,
+      "loss": 1.952,
+      "step": 11947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.754595497597667e-05,
+      "loss": 1.9431,
+      "step": 11948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.7545561534716938e-05,
+      "loss": 1.8034,
+      "step": 11949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.75451680663327e-05,
+      "loss": 1.5758,
+      "step": 11950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7544774570825367e-05,
+      "loss": 2.0068,
+      "step": 11951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7544381048196353e-05,
+      "loss": 1.8243,
+      "step": 11952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7543987498447076e-05,
+      "loss": 1.8603,
+      "step": 11953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.754359392157894e-05,
+      "loss": 1.8825,
+      "step": 11954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.7543200317593376e-05,
+      "loss": 1.5328,
+      "step": 11955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7542806686491788e-05,
+      "loss": 1.8836,
+      "step": 11956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7542413028275595e-05,
+      "loss": 1.9562,
+      "step": 11957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.754201934294621e-05,
+      "loss": 1.96,
+      "step": 11958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7541625630505048e-05,
+      "loss": 1.8806,
+      "step": 11959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.754123189095353e-05,
+      "loss": 1.9608,
+      "step": 11960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.754083812429306e-05,
+      "loss": 1.981,
+      "step": 11961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7540444330525062e-05,
+      "loss": 1.718,
+      "step": 11962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.754005050965095e-05,
+      "loss": 1.6225,
+      "step": 11963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7539656661672144e-05,
+      "loss": 2.2565,
+      "step": 11964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.753926278659005e-05,
+      "loss": 1.9629,
+      "step": 11965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7538868884406092e-05,
+      "loss": 1.6599,
+      "step": 11966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7538474955121684e-05,
+      "loss": 1.871,
+      "step": 11967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7538080998738238e-05,
+      "loss": 1.7895,
+      "step": 11968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7537687015257175e-05,
+      "loss": 2.0435,
+      "step": 11969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7537293004679914e-05,
+      "loss": 2.3726,
+      "step": 11970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7536898967007862e-05,
+      "loss": 1.8359,
+      "step": 11971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7536504902242444e-05,
+      "loss": 1.8941,
+      "step": 11972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.753611081038507e-05,
+      "loss": 1.8075,
+      "step": 11973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.753571669143716e-05,
+      "loss": 1.7605,
+      "step": 11974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7535322545400132e-05,
+      "loss": 1.9984,
+      "step": 11975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7534928372275402e-05,
+      "loss": 1.7626,
+      "step": 11976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.753453417206439e-05,
+      "loss": 1.9366,
+      "step": 11977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7534139944768503e-05,
+      "loss": 1.9213,
+      "step": 11978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7533745690389167e-05,
+      "loss": 1.9123,
+      "step": 11979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7533351408927798e-05,
+      "loss": 2.2289,
+      "step": 11980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.753295710038581e-05,
+      "loss": 1.8101,
+      "step": 11981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7532562764764623e-05,
+      "loss": 1.7814,
+      "step": 11982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7532168402065656e-05,
+      "loss": 1.8502,
+      "step": 11983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7531774012290322e-05,
+      "loss": 1.7276,
+      "step": 11984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7531379595440044e-05,
+      "loss": 2.0731,
+      "step": 11985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7530985151516238e-05,
+      "loss": 1.8357,
+      "step": 11986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.753059068052032e-05,
+      "loss": 1.9099,
+      "step": 11987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7530196182453712e-05,
+      "loss": 1.7298,
+      "step": 11988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.7529801657317827e-05,
+      "loss": 1.7535,
+      "step": 11989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7529407105114086e-05,
+      "loss": 1.9886,
+      "step": 11990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.752901252584391e-05,
+      "loss": 1.7698,
+      "step": 11991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.752861791950871e-05,
+      "loss": 1.553,
+      "step": 11992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.752822328610991e-05,
+      "loss": 1.7552,
+      "step": 11993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.752782862564893e-05,
+      "loss": 1.9089,
+      "step": 11994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7527433938127188e-05,
+      "loss": 1.8728,
+      "step": 11995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7527039223546096e-05,
+      "loss": 1.9163,
+      "step": 11996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.752664448190708e-05,
+      "loss": 1.7102,
+      "step": 11997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.752624971321156e-05,
+      "loss": 2.0849,
+      "step": 11998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.752585491746095e-05,
+      "loss": 2.1584,
+      "step": 11999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.752546009465667e-05,
+      "loss": 2.0579,
+      "step": 12000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7525065244800142e-05,
+      "loss": 1.7757,
+      "step": 12001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7524670367892782e-05,
+      "loss": 1.8306,
+      "step": 12002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7524275463936014e-05,
+      "loss": 1.8713,
+      "step": 12003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7523880532931255e-05,
+      "loss": 1.4805,
+      "step": 12004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7523485574879925e-05,
+      "loss": 2.0042,
+      "step": 12005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7523090589783445e-05,
+      "loss": 1.8765,
+      "step": 12006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.752269557764323e-05,
+      "loss": 1.8857,
+      "step": 12007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7522300538460705e-05,
+      "loss": 1.9608,
+      "step": 12008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.752190547223729e-05,
+      "loss": 1.8571,
+      "step": 12009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.75215103789744e-05,
+      "loss": 1.7088,
+      "step": 12010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.752111525867346e-05,
+      "loss": 1.8211,
+      "step": 12011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7520720111335892e-05,
+      "loss": 2.0222,
+      "step": 12012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7520324936963114e-05,
+      "loss": 1.5412,
+      "step": 12013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7519929735556544e-05,
+      "loss": 1.9638,
+      "step": 12014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7519534507117603e-05,
+      "loss": 1.7729,
+      "step": 12015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.751913925164772e-05,
+      "loss": 1.6054,
+      "step": 12016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7518743969148305e-05,
+      "loss": 1.6516,
+      "step": 12017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7518348659620785e-05,
+      "loss": 1.8345,
+      "step": 12018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7517953323066578e-05,
+      "loss": 1.881,
+      "step": 12019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7517557959487107e-05,
+      "loss": 1.8718,
+      "step": 12020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7517162568883795e-05,
+      "loss": 1.6482,
+      "step": 12021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.751676715125806e-05,
+      "loss": 2.0245,
+      "step": 12022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7516371706611325e-05,
+      "loss": 1.9977,
+      "step": 12023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.7515976234945012e-05,
+      "loss": 1.981,
+      "step": 12024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.751558073626054e-05,
+      "loss": 1.7197,
+      "step": 12025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7515185210559336e-05,
+      "loss": 1.836,
+      "step": 12026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.7514789657842816e-05,
+      "loss": 2.1308,
+      "step": 12027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7514394078112406e-05,
+      "loss": 1.7335,
+      "step": 12028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7513998471369525e-05,
+      "loss": 1.8502,
+      "step": 12029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7513602837615594e-05,
+      "loss": 1.7194,
+      "step": 12030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7513207176852042e-05,
+      "loss": 1.7277,
+      "step": 12031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7512811489080286e-05,
+      "loss": 1.8663,
+      "step": 12032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7512415774301748e-05,
+      "loss": 1.8456,
+      "step": 12033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7512020032517855e-05,
+      "loss": 1.8551,
+      "step": 12034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7511624263730023e-05,
+      "loss": 1.9315,
+      "step": 12035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.751122846793968e-05,
+      "loss": 1.9965,
+      "step": 12036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7510832645148244e-05,
+      "loss": 1.8395,
+      "step": 12037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7510436795357147e-05,
+      "loss": 2.1859,
+      "step": 12038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.75100409185678e-05,
+      "loss": 1.9492,
+      "step": 12039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7509645014781635e-05,
+      "loss": 2.1351,
+      "step": 12040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7509249084000073e-05,
+      "loss": 1.7053,
+      "step": 12041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7508853126224534e-05,
+      "loss": 1.8646,
+      "step": 12042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7508457141456446e-05,
+      "loss": 1.675,
+      "step": 12043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.7508061129697227e-05,
+      "loss": 2.1299,
+      "step": 12044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7507665090948305e-05,
+      "loss": 1.9955,
+      "step": 12045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7507269025211105e-05,
+      "loss": 1.7559,
+      "step": 12046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7506872932487045e-05,
+      "loss": 1.9492,
+      "step": 12047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7506476812777556e-05,
+      "loss": 2.0392,
+      "step": 12048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.7506080666084054e-05,
+      "loss": 1.897,
+      "step": 12049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.750568449240797e-05,
+      "loss": 1.5514,
+      "step": 12050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7505288291750724e-05,
+      "loss": 1.9778,
+      "step": 12051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.750489206411374e-05,
+      "loss": 2.0256,
+      "step": 12052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7504495809498445e-05,
+      "loss": 2.1407,
+      "step": 12053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7504099527906265e-05,
+      "loss": 1.8618,
+      "step": 12054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.375,
+      "learning_rate": 1.750370321933862e-05,
+      "loss": 1.971,
+      "step": 12055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7503306883796937e-05,
+      "loss": 2.211,
+      "step": 12056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7502910521282638e-05,
+      "loss": 2.0396,
+      "step": 12057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7502514131797153e-05,
+      "loss": 1.7234,
+      "step": 12058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7502117715341902e-05,
+      "loss": 1.9736,
+      "step": 12059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.750172127191831e-05,
+      "loss": 1.5646,
+      "step": 12060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.750132480152781e-05,
+      "loss": 1.5305,
+      "step": 12061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.750092830417182e-05,
+      "loss": 1.9315,
+      "step": 12062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7500531779851763e-05,
+      "loss": 1.6311,
+      "step": 12063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7500135228569067e-05,
+      "loss": 1.1316,
+      "step": 12064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7499738650325163e-05,
+      "loss": 1.7394,
+      "step": 12065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.749934204512147e-05,
+      "loss": 1.9417,
+      "step": 12066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7498945412959417e-05,
+      "loss": 2.0641,
+      "step": 12067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.749854875384043e-05,
+      "loss": 1.8204,
+      "step": 12068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7498152067765932e-05,
+      "loss": 1.8592,
+      "step": 12069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7497755354737348e-05,
+      "loss": 1.9938,
+      "step": 12070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.749735861475611e-05,
+      "loss": 3.4598,
+      "step": 12071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7496961847823638e-05,
+      "loss": 2.1022,
+      "step": 12072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7496565053941364e-05,
+      "loss": 2.2649,
+      "step": 12073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7496168233110714e-05,
+      "loss": 2.1805,
+      "step": 12074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7495771385333106e-05,
+      "loss": 1.8646,
+      "step": 12075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7495374510609978e-05,
+      "loss": 1.6034,
+      "step": 12076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7494977608942748e-05,
+      "loss": 2.0165,
+      "step": 12077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7494580680332847e-05,
+      "loss": 2.0826,
+      "step": 12078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7494183724781698e-05,
+      "loss": 2.1135,
+      "step": 12079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7493786742290734e-05,
+      "loss": 1.83,
+      "step": 12080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.749338973286138e-05,
+      "loss": 2.2509,
+      "step": 12081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.749299269649506e-05,
+      "loss": 1.7586,
+      "step": 12082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.74925956331932e-05,
+      "loss": 1.7846,
+      "step": 12083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7492198542957236e-05,
+      "loss": 1.8693,
+      "step": 12084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7491801425788586e-05,
+      "loss": 2.095,
+      "step": 12085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7491404281688685e-05,
+      "loss": 1.7815,
+      "step": 12086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7491007110658954e-05,
+      "loss": 2.1648,
+      "step": 12087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7490609912700826e-05,
+      "loss": 1.7066,
+      "step": 12088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7490212687815727e-05,
+      "loss": 1.776,
+      "step": 12089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7489815436005088e-05,
+      "loss": 1.8183,
+      "step": 12090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.748941815727033e-05,
+      "loss": 1.6789,
+      "step": 12091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7489020851612884e-05,
+      "loss": 1.9225,
+      "step": 12092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.748862351903418e-05,
+      "loss": 2.1145,
+      "step": 12093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7488226159535648e-05,
+      "loss": 1.9126,
+      "step": 12094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.748782877311871e-05,
+      "loss": 2.2472,
+      "step": 12095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7487431359784803e-05,
+      "loss": 2.2223,
+      "step": 12096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.7487033919535346e-05,
+      "loss": 2.1088,
+      "step": 12097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7486636452371777e-05,
+      "loss": 1.9617,
+      "step": 12098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7486238958295515e-05,
+      "loss": 1.9558,
+      "step": 12099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7485841437308e-05,
+      "loss": 1.9735,
+      "step": 12100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.748544388941065e-05,
+      "loss": 1.9206,
+      "step": 12101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7485046314604905e-05,
+      "loss": 2.0048,
+      "step": 12102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7484648712892188e-05,
+      "loss": 1.5855,
+      "step": 12103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7484251084273928e-05,
+      "loss": 2.0856,
+      "step": 12104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7483853428751552e-05,
+      "loss": 2.0217,
+      "step": 12105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7483455746326496e-05,
+      "loss": 1.8645,
+      "step": 12106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7483058037000185e-05,
+      "loss": 2.1696,
+      "step": 12107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7482660300774052e-05,
+      "loss": 1.7255,
+      "step": 12108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.748226253764952e-05,
+      "loss": 1.6383,
+      "step": 12109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.748186474762803e-05,
+      "loss": 2.0024,
+      "step": 12110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.7481466930711006e-05,
+      "loss": 2.2571,
+      "step": 12111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7481069086899874e-05,
+      "loss": 1.902,
+      "step": 12112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7480671216196067e-05,
+      "loss": 2.0703,
+      "step": 12113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7480273318601017e-05,
+      "loss": 1.915,
+      "step": 12114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7479875394116153e-05,
+      "loss": 1.7237,
+      "step": 12115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7479477442742907e-05,
+      "loss": 2.0921,
+      "step": 12116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.747907946448271e-05,
+      "loss": 2.0874,
+      "step": 12117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.747868145933699e-05,
+      "loss": 2.0657,
+      "step": 12118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.747828342730718e-05,
+      "loss": 1.6148,
+      "step": 12119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7477885368394703e-05,
+      "loss": 1.7167,
+      "step": 12120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7477487282601004e-05,
+      "loss": 1.6786,
+      "step": 12121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7477089169927504e-05,
+      "loss": 1.9816,
+      "step": 12122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.7476691030375635e-05,
+      "loss": 1.8859,
+      "step": 12123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7476292863946832e-05,
+      "loss": 1.786,
+      "step": 12124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7475894670642527e-05,
+      "loss": 1.7728,
+      "step": 12125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7475496450464144e-05,
+      "loss": 1.9841,
+      "step": 12126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7475098203413122e-05,
+      "loss": 1.8425,
+      "step": 12127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7474699929490888e-05,
+      "loss": 1.8019,
+      "step": 12128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7474301628698874e-05,
+      "loss": 1.8627,
+      "step": 12129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.747390330103852e-05,
+      "loss": 2.0144,
+      "step": 12130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7473504946511243e-05,
+      "loss": 2.2299,
+      "step": 12131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.747310656511849e-05,
+      "loss": 1.9225,
+      "step": 12132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.747270815686168e-05,
+      "loss": 1.6518,
+      "step": 12133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7472309721742253e-05,
+      "loss": 1.9112,
+      "step": 12134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.747191125976164e-05,
+      "loss": 1.9432,
+      "step": 12135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7471512770921274e-05,
+      "loss": 1.9216,
+      "step": 12136
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.747111425522259e-05,
+      "loss": 1.4512,
+      "step": 12137
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7470715712667013e-05,
+      "loss": 1.6272,
+      "step": 12138
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7470317143255983e-05,
+      "loss": 2.0886,
+      "step": 12139
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7469918546990926e-05,
+      "loss": 1.9171,
+      "step": 12140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7469519923873277e-05,
+      "loss": 1.7857,
+      "step": 12141
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.746912127390447e-05,
+      "loss": 2.0463,
+      "step": 12142
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7468722597085947e-05,
+      "loss": 2.0153,
+      "step": 12143
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.7468323893419124e-05,
+      "loss": 1.7979,
+      "step": 12144
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7467925162905447e-05,
+      "loss": 1.9631,
+      "step": 12145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7467526405546344e-05,
+      "loss": 1.8755,
+      "step": 12146
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7467127621343246e-05,
+      "loss": 2.0073,
+      "step": 12147
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.7466728810297596e-05,
+      "loss": 3.2437,
+      "step": 12148
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7466329972410817e-05,
+      "loss": 1.6241,
+      "step": 12149
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7465931107684352e-05,
+      "loss": 1.672,
+      "step": 12150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7465532216119628e-05,
+      "loss": 1.8177,
+      "step": 12151
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.746513329771808e-05,
+      "loss": 2.4808,
+      "step": 12152
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7464734352481145e-05,
+      "loss": 1.6158,
+      "step": 12153
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7464335380410256e-05,
+      "loss": 1.7818,
+      "step": 12154
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7463936381506844e-05,
+      "loss": 1.6831,
+      "step": 12155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7463537355772348e-05,
+      "loss": 1.7092,
+      "step": 12156
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.74631383032082e-05,
+      "loss": 1.9523,
+      "step": 12157
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7462739223815834e-05,
+      "loss": 1.7772,
+      "step": 12158
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7462340117596688e-05,
+      "loss": 2.0958,
+      "step": 12159
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.746194098455219e-05,
+      "loss": 1.7849,
+      "step": 12160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.746154182468378e-05,
+      "loss": 2.2124,
+      "step": 12161
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7461142637992893e-05,
+      "loss": 1.9138,
+      "step": 12162
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7460743424480964e-05,
+      "loss": 1.984,
+      "step": 12163
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.746034418414942e-05,
+      "loss": 1.7222,
+      "step": 12164
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.745994491699971e-05,
+      "loss": 1.5931,
+      "step": 12165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7459545623033262e-05,
+      "loss": 1.8946,
+      "step": 12166
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.745914630225151e-05,
+      "loss": 1.8889,
+      "step": 12167
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.745874695465589e-05,
+      "loss": 1.8463,
+      "step": 12168
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7458347580247842e-05,
+      "loss": 1.798,
+      "step": 12169
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7457948179028795e-05,
+      "loss": 1.5597,
+      "step": 12170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.745754875100019e-05,
+      "loss": 2.3712,
+      "step": 12171
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7457149296163457e-05,
+      "loss": 1.6856,
+      "step": 12172
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.745674981452004e-05,
+      "loss": 2.0041,
+      "step": 12173
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.745635030607137e-05,
+      "loss": 2.058,
+      "step": 12174
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7455950770818885e-05,
+      "loss": 2.0887,
+      "step": 12175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7455551208764018e-05,
+      "loss": 1.9743,
+      "step": 12176
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.745515161990821e-05,
+      "loss": 1.7473,
+      "step": 12177
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.745475200425289e-05,
+      "loss": 1.6137,
+      "step": 12178
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7454352361799504e-05,
+      "loss": 1.8353,
+      "step": 12179
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7453952692549484e-05,
+      "loss": 1.8583,
+      "step": 12180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7453552996504264e-05,
+      "loss": 1.5727,
+      "step": 12181
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.745315327366529e-05,
+      "loss": 1.9449,
+      "step": 12182
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7452753524033987e-05,
+      "loss": 2.0005,
+      "step": 12183
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7452353747611795e-05,
+      "loss": 1.6082,
+      "step": 12184
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.745195394440016e-05,
+      "loss": 1.7968,
+      "step": 12185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7451554114400512e-05,
+      "loss": 1.8595,
+      "step": 12186
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7451154257614287e-05,
+      "loss": 1.847,
+      "step": 12187
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7450754374042926e-05,
+      "loss": 2.2002,
+      "step": 12188
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.7450354463687864e-05,
+      "loss": 1.7704,
+      "step": 12189
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.744995452655054e-05,
+      "loss": 1.9213,
+      "step": 12190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.744955456263239e-05,
+      "loss": 2.0664,
+      "step": 12191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7449154571934857e-05,
+      "loss": 1.8576,
+      "step": 12192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.744875455445937e-05,
+      "loss": 1.7435,
+      "step": 12193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7448354510207373e-05,
+      "loss": 1.5659,
+      "step": 12194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.744795443918031e-05,
+      "loss": 1.914,
+      "step": 12195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7447554341379604e-05,
+      "loss": 1.892,
+      "step": 12196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7447154216806704e-05,
+      "loss": 2.138,
+      "step": 12197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7446754065463048e-05,
+      "loss": 1.6095,
+      "step": 12198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.7446353887350067e-05,
+      "loss": 1.8162,
+      "step": 12199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.744595368246921e-05,
+      "loss": 2.1536,
+      "step": 12200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7445553450821908e-05,
+      "loss": 2.0062,
+      "step": 12201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7445153192409604e-05,
+      "loss": 2.3953,
+      "step": 12202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7444752907233732e-05,
+      "loss": 2.0565,
+      "step": 12203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.744435259529574e-05,
+      "loss": 1.8822,
+      "step": 12204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.744395225659705e-05,
+      "loss": 1.8612,
+      "step": 12205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7443551891139118e-05,
+      "loss": 1.8942,
+      "step": 12206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.744315149892338e-05,
+      "loss": 1.9619,
+      "step": 12207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.744275107995127e-05,
+      "loss": 2.3522,
+      "step": 12208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.744235063422423e-05,
+      "loss": 1.7604,
+      "step": 12209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7441950161743703e-05,
+      "loss": 2.0847,
+      "step": 12210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7441549662511122e-05,
+      "loss": 2.2051,
+      "step": 12211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7441149136527927e-05,
+      "loss": 2.0429,
+      "step": 12212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7440748583795563e-05,
+      "loss": 1.9637,
+      "step": 12213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.744034800431547e-05,
+      "loss": 2.1959,
+      "step": 12214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7439947398089084e-05,
+      "loss": 1.9388,
+      "step": 12215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7439546765117845e-05,
+      "loss": 2.0806,
+      "step": 12216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.74391461054032e-05,
+      "loss": 1.8502,
+      "step": 12217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7438745418946578e-05,
+      "loss": 2.2896,
+      "step": 12218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.743834470574943e-05,
+      "loss": 1.7869,
+      "step": 12219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7437943965813186e-05,
+      "loss": 2.1504,
+      "step": 12220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7437543199139297e-05,
+      "loss": 1.9272,
+      "step": 12221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7437142405729196e-05,
+      "loss": 1.7793,
+      "step": 12222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.743674158558433e-05,
+      "loss": 1.9031,
+      "step": 12223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7436340738706138e-05,
+      "loss": 1.9598,
+      "step": 12224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7435939865096054e-05,
+      "loss": 1.8632,
+      "step": 12225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.743553896475553e-05,
+      "loss": 1.8578,
+      "step": 12226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7435138037686e-05,
+      "loss": 2.1255,
+      "step": 12227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7434737083888905e-05,
+      "loss": 1.8079,
+      "step": 12228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7434336103365696e-05,
+      "loss": 1.9438,
+      "step": 12229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7433935096117797e-05,
+      "loss": 2.1267,
+      "step": 12230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7433534062146663e-05,
+      "loss": 1.7679,
+      "step": 12231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7433133001453733e-05,
+      "loss": 1.9355,
+      "step": 12232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.743273191404045e-05,
+      "loss": 1.689,
+      "step": 12233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.743233079990825e-05,
+      "loss": 2.174,
+      "step": 12234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7431929659058583e-05,
+      "loss": 1.9424,
+      "step": 12235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.743152849149288e-05,
+      "loss": 1.8554,
+      "step": 12236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7431127297212593e-05,
+      "loss": 1.972,
+      "step": 12237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7430726076219162e-05,
+      "loss": 1.6888,
+      "step": 12238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7430324828514026e-05,
+      "loss": 1.7428,
+      "step": 12239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.742992355409863e-05,
+      "loss": 1.8519,
+      "step": 12240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7429522252974416e-05,
+      "loss": 1.8623,
+      "step": 12241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7429120925142826e-05,
+      "loss": 1.946,
+      "step": 12242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7428719570605303e-05,
+      "loss": 1.8922,
+      "step": 12243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7428318189363292e-05,
+      "loss": 1.906,
+      "step": 12244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.742791678141823e-05,
+      "loss": 1.7158,
+      "step": 12245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7427515346771568e-05,
+      "loss": 2.0263,
+      "step": 12246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.7427113885424743e-05,
+      "loss": 1.9809,
+      "step": 12247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.74267123973792e-05,
+      "loss": 1.8468,
+      "step": 12248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7426310882636385e-05,
+      "loss": 2.0312,
+      "step": 12249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7425909341197736e-05,
+      "loss": 1.5948,
+      "step": 12250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7425507773064698e-05,
+      "loss": 2.2387,
+      "step": 12251
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7425106178238717e-05,
+      "loss": 1.6092,
+      "step": 12252
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7424704556721238e-05,
+      "loss": 1.4095,
+      "step": 12253
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7424302908513696e-05,
+      "loss": 1.8522,
+      "step": 12254
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7423901233617545e-05,
+      "loss": 2.1104,
+      "step": 12255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7423499532034226e-05,
+      "loss": 1.7718,
+      "step": 12256
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7423097803765177e-05,
+      "loss": 2.1217,
+      "step": 12257
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.742269604881185e-05,
+      "loss": 2.1529,
+      "step": 12258
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7422294267175687e-05,
+      "loss": 1.8865,
+      "step": 12259
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.742189245885813e-05,
+      "loss": 1.8098,
+      "step": 12260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7421490623860625e-05,
+      "loss": 1.7649,
+      "step": 12261
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7421088762184616e-05,
+      "loss": 2.001,
+      "step": 12262
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7420686873831548e-05,
+      "loss": 1.8137,
+      "step": 12263
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7420284958802868e-05,
+      "loss": 1.9076,
+      "step": 12264
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7419883017100012e-05,
+      "loss": 1.5791,
+      "step": 12265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7419481048724435e-05,
+      "loss": 1.9599,
+      "step": 12266
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7419079053677582e-05,
+      "loss": 2.0745,
+      "step": 12267
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.741867703196089e-05,
+      "loss": 1.9024,
+      "step": 12268
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7418274983575808e-05,
+      "loss": 2.2246,
+      "step": 12269
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7417872908523787e-05,
+      "loss": 1.7748,
+      "step": 12270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7417470806806263e-05,
+      "loss": 1.7937,
+      "step": 12271
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7417068678424685e-05,
+      "loss": 1.8567,
+      "step": 12272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7416666523380502e-05,
+      "loss": 2.0272,
+      "step": 12273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7416264341675157e-05,
+      "loss": 1.9021,
+      "step": 12274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7415862133310094e-05,
+      "loss": 2.157,
+      "step": 12275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.741545989828676e-05,
+      "loss": 2.0461,
+      "step": 12276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.74150576366066e-05,
+      "loss": 1.8666,
+      "step": 12277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7414655348271065e-05,
+      "loss": 2.1279,
+      "step": 12278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7414253033281597e-05,
+      "loss": 1.7789,
+      "step": 12279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7413850691639636e-05,
+      "loss": 1.9268,
+      "step": 12280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7413448323346645e-05,
+      "loss": 1.6287,
+      "step": 12281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7413045928404052e-05,
+      "loss": 1.858,
+      "step": 12282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.741264350681332e-05,
+      "loss": 1.8241,
+      "step": 12283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7412241058575884e-05,
+      "loss": 1.7008,
+      "step": 12284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.741183858369319e-05,
+      "loss": 1.773,
+      "step": 12285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7411436082166698e-05,
+      "loss": 1.8711,
+      "step": 12286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7411033553997837e-05,
+      "loss": 1.5047,
+      "step": 12287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7410630999188072e-05,
+      "loss": 2.0424,
+      "step": 12288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7410228417738836e-05,
+      "loss": 1.8136,
+      "step": 12289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.7409825809651583e-05,
+      "loss": 2.1957,
+      "step": 12290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7409423174927753e-05,
+      "loss": 1.8402,
+      "step": 12291
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7409020513568805e-05,
+      "loss": 1.8074,
+      "step": 12292
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.740861782557618e-05,
+      "loss": 1.6473,
+      "step": 12293
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7408215110951326e-05,
+      "loss": 2.1672,
+      "step": 12294
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7407812369695688e-05,
+      "loss": 2.1493,
+      "step": 12295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.740740960181072e-05,
+      "loss": 1.9881,
+      "step": 12296
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.740700680729786e-05,
+      "loss": 1.771,
+      "step": 12297
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.7406603986158568e-05,
+      "loss": 1.8487,
+      "step": 12298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7406201138394286e-05,
+      "loss": 1.9188,
+      "step": 12299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7405798264006458e-05,
+      "loss": 1.9514,
+      "step": 12300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7405395362996543e-05,
+      "loss": 2.6534,
+      "step": 12301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.740499243536598e-05,
+      "loss": 1.8257,
+      "step": 12302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7404589481116218e-05,
+      "loss": 1.8222,
+      "step": 12303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.740418650024871e-05,
+      "loss": 1.5851,
+      "step": 12304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7403783492764905e-05,
+      "loss": 1.8892,
+      "step": 12305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7403380458666246e-05,
+      "loss": 1.8547,
+      "step": 12306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7402977397954184e-05,
+      "loss": 1.6472,
+      "step": 12307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.740257431063017e-05,
+      "loss": 1.865,
+      "step": 12308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7402171196695655e-05,
+      "loss": 2.007,
+      "step": 12309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7401768056152083e-05,
+      "loss": 1.7158,
+      "step": 12310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7401364889000906e-05,
+      "loss": 1.9715,
+      "step": 12311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.740096169524357e-05,
+      "loss": 2.156,
+      "step": 12312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7400558474881533e-05,
+      "loss": 1.4773,
+      "step": 12313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7400155227916234e-05,
+      "loss": 2.0279,
+      "step": 12314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7399751954349126e-05,
+      "loss": 2.0471,
+      "step": 12315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7399348654181664e-05,
+      "loss": 2.2254,
+      "step": 12316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.739894532741529e-05,
+      "loss": 1.862,
+      "step": 12317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.739854197405146e-05,
+      "loss": 1.8086,
+      "step": 12318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7398138594091615e-05,
+      "loss": 1.9101,
+      "step": 12319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7397735187537218e-05,
+      "loss": 2.1642,
+      "step": 12320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.739733175438971e-05,
+      "loss": 1.7043,
+      "step": 12321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7396928294650544e-05,
+      "loss": 1.6759,
+      "step": 12322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.739652480832117e-05,
+      "loss": 2.043,
+      "step": 12323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.739612129540304e-05,
+      "loss": 1.8985,
+      "step": 12324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.73957177558976e-05,
+      "loss": 2.3511,
+      "step": 12325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.7395314189806308e-05,
+      "loss": 1.9271,
+      "step": 12326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7394910597130607e-05,
+      "loss": 1.952,
+      "step": 12327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7394506977871952e-05,
+      "loss": 1.8478,
+      "step": 12328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7394103332031794e-05,
+      "loss": 1.7458,
+      "step": 12329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7393699659611583e-05,
+      "loss": 2.1165,
+      "step": 12330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7393295960612772e-05,
+      "loss": 1.8361,
+      "step": 12331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7392892235036805e-05,
+      "loss": 2.1651,
+      "step": 12332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.7392488482885142e-05,
+      "loss": 1.9944,
+      "step": 12333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7392084704159233e-05,
+      "loss": 1.8843,
+      "step": 12334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7391680898860525e-05,
+      "loss": 1.9914,
+      "step": 12335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.739127706699047e-05,
+      "loss": 1.9555,
+      "step": 12336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.739087320855053e-05,
+      "loss": 1.652,
+      "step": 12337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.739046932354214e-05,
+      "loss": 2.0088,
+      "step": 12338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7390065411966765e-05,
+      "loss": 2.1954,
+      "step": 12339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7389661473825854e-05,
+      "loss": 1.7466,
+      "step": 12340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7389257509120858e-05,
+      "loss": 1.6151,
+      "step": 12341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7388853517853223e-05,
+      "loss": 1.7914,
+      "step": 12342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.738844950002441e-05,
+      "loss": 1.8328,
+      "step": 12343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7388045455635868e-05,
+      "loss": 2.2341,
+      "step": 12344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.738764138468905e-05,
+      "loss": 1.8629,
+      "step": 12345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.738723728718541e-05,
+      "loss": 1.7859,
+      "step": 12346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7386833163126397e-05,
+      "loss": 1.7817,
+      "step": 12347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.738642901251347e-05,
+      "loss": 1.9614,
+      "step": 12348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7386024835348072e-05,
+      "loss": 1.9309,
+      "step": 12349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7385620631631662e-05,
+      "loss": 1.783,
+      "step": 12350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7385216401365696e-05,
+      "loss": 1.6862,
+      "step": 12351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7384812144551623e-05,
+      "loss": 1.8618,
+      "step": 12352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7384407861190895e-05,
+      "loss": 1.8845,
+      "step": 12353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.738400355128497e-05,
+      "loss": 1.704,
+      "step": 12354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7383599214835295e-05,
+      "loss": 1.933,
+      "step": 12355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7383194851843325e-05,
+      "loss": 1.8243,
+      "step": 12356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.738279046231052e-05,
+      "loss": 1.8947,
+      "step": 12357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.738238604623833e-05,
+      "loss": 1.7615,
+      "step": 12358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7381981603628203e-05,
+      "loss": 1.8689,
+      "step": 12359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7381577134481602e-05,
+      "loss": 1.9225,
+      "step": 12360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7381172638799976e-05,
+      "loss": 2.2447,
+      "step": 12361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.738076811658478e-05,
+      "loss": 1.776,
+      "step": 12362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7380363567837467e-05,
+      "loss": 1.6759,
+      "step": 12363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7379958992559494e-05,
+      "loss": 2.023,
+      "step": 12364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7379554390752314e-05,
+      "loss": 1.6854,
+      "step": 12365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7379149762417376e-05,
+      "loss": 2.0794,
+      "step": 12366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7378745107556145e-05,
+      "loss": 1.613,
+      "step": 12367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7378340426170074e-05,
+      "loss": 1.7147,
+      "step": 12368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.737793571826061e-05,
+      "loss": 1.9236,
+      "step": 12369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.737753098382921e-05,
+      "loss": 1.8837,
+      "step": 12370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.737712622287733e-05,
+      "loss": 2.1056,
+      "step": 12371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.737672143540643e-05,
+      "loss": 1.8018,
+      "step": 12372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7376316621417958e-05,
+      "loss": 2.5856,
+      "step": 12373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.737591178091337e-05,
+      "loss": 1.9338,
+      "step": 12374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.737550691389413e-05,
+      "loss": 1.9173,
+      "step": 12375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7375102020361684e-05,
+      "loss": 2.0929,
+      "step": 12376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.737469710031749e-05,
+      "loss": 1.7742,
+      "step": 12377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7374292153763e-05,
+      "loss": 2.075,
+      "step": 12378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.737388718069968e-05,
+      "loss": 1.7384,
+      "step": 12379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7373482181128976e-05,
+      "loss": 1.8642,
+      "step": 12380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7373077155052347e-05,
+      "loss": 2.0429,
+      "step": 12381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.737267210247125e-05,
+      "loss": 1.9429,
+      "step": 12382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7372267023387145e-05,
+      "loss": 2.0539,
+      "step": 12383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7371861917801478e-05,
+      "loss": 1.6642,
+      "step": 12384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7371456785715715e-05,
+      "loss": 1.9515,
+      "step": 12385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7371051627131305e-05,
+      "loss": 1.6583,
+      "step": 12386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.737064644204971e-05,
+      "loss": 1.812,
+      "step": 12387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.737024123047238e-05,
+      "loss": 1.6016,
+      "step": 12388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.736983599240078e-05,
+      "loss": 1.7403,
+      "step": 12389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.736943072783636e-05,
+      "loss": 1.8034,
+      "step": 12390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.736902543678058e-05,
+      "loss": 1.7921,
+      "step": 12391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7368620119234896e-05,
+      "loss": 2.2159,
+      "step": 12392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7368214775200765e-05,
+      "loss": 1.6238,
+      "step": 12393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7367809404679645e-05,
+      "loss": 1.5616,
+      "step": 12394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7367404007672994e-05,
+      "loss": 2.1646,
+      "step": 12395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7366998584182265e-05,
+      "loss": 1.8528,
+      "step": 12396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.736659313420892e-05,
+      "loss": 1.8325,
+      "step": 12397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7366187657754414e-05,
+      "loss": 1.8956,
+      "step": 12398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7365782154820208e-05,
+      "loss": 2.1326,
+      "step": 12399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7365376625407755e-05,
+      "loss": 1.9864,
+      "step": 12400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7364971069518516e-05,
+      "loss": 1.9923,
+      "step": 12401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7364565487153944e-05,
+      "loss": 1.6669,
+      "step": 12402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7364159878315504e-05,
+      "loss": 2.0007,
+      "step": 12403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7363754243004654e-05,
+      "loss": 1.6414,
+      "step": 12404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7363348581222843e-05,
+      "loss": 1.9602,
+      "step": 12405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.736294289297154e-05,
+      "loss": 1.987,
+      "step": 12406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7362537178252195e-05,
+      "loss": 2.0852,
+      "step": 12407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.736213143706627e-05,
+      "loss": 2.0689,
+      "step": 12408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7361725669415226e-05,
+      "loss": 1.7491,
+      "step": 12409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7361319875300518e-05,
+      "loss": 1.8633,
+      "step": 12410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.7360914054723607e-05,
+      "loss": 1.9048,
+      "step": 12411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7360508207685953e-05,
+      "loss": 1.9566,
+      "step": 12412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7360102334189005e-05,
+      "loss": 2.0481,
+      "step": 12413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7359696434234234e-05,
+      "loss": 1.8411,
+      "step": 12414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7359290507823097e-05,
+      "loss": 2.0008,
+      "step": 12415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.735888455495705e-05,
+      "loss": 1.7228,
+      "step": 12416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7358478575637553e-05,
+      "loss": 1.6777,
+      "step": 12417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7358072569866063e-05,
+      "loss": 2.1124,
+      "step": 12418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7357666537644045e-05,
+      "loss": 2.0766,
+      "step": 12419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.735726047897296e-05,
+      "loss": 1.9474,
+      "step": 12420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7356854393854255e-05,
+      "loss": 1.8946,
+      "step": 12421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7356448282289403e-05,
+      "loss": 1.5257,
+      "step": 12422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7356042144279858e-05,
+      "loss": 2.2216,
+      "step": 12423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7355635979827082e-05,
+      "loss": 1.7725,
+      "step": 12424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7355229788932534e-05,
+      "loss": 1.6567,
+      "step": 12425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7354823571597675e-05,
+      "loss": 1.7885,
+      "step": 12426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7354417327823964e-05,
+      "loss": 2.1228,
+      "step": 12427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.735401105761286e-05,
+      "loss": 1.7779,
+      "step": 12428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7353604760965828e-05,
+      "loss": 2.0983,
+      "step": 12429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7353198437884324e-05,
+      "loss": 1.966,
+      "step": 12430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.735279208836981e-05,
+      "loss": 1.6608,
+      "step": 12431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.735238571242375e-05,
+      "loss": 1.7472,
+      "step": 12432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7351979310047603e-05,
+      "loss": 1.7317,
+      "step": 12433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7351572881242828e-05,
+      "loss": 1.8129,
+      "step": 12434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7351166426010885e-05,
+      "loss": 1.9314,
+      "step": 12435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7350759944353236e-05,
+      "loss": 1.8364,
+      "step": 12436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7350353436271347e-05,
+      "loss": 1.7122,
+      "step": 12437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7349946901766675e-05,
+      "loss": 1.9411,
+      "step": 12438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7349540340840684e-05,
+      "loss": 1.7816,
+      "step": 12439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7349133753494828e-05,
+      "loss": 1.922,
+      "step": 12440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.734872713973058e-05,
+      "loss": 1.6794,
+      "step": 12441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.7348320499549393e-05,
+      "loss": 1.9143,
+      "step": 12442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.734791383295273e-05,
+      "loss": 2.0967,
+      "step": 12443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7347507139942056e-05,
+      "loss": 2.1063,
+      "step": 12444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7347100420518833e-05,
+      "loss": 1.8607,
+      "step": 12445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7346693674684523e-05,
+      "loss": 1.9033,
+      "step": 12446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7346286902440583e-05,
+      "loss": 1.986,
+      "step": 12447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7345880103788483e-05,
+      "loss": 1.6184,
+      "step": 12448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7345473278729677e-05,
+      "loss": 1.6811,
+      "step": 12449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7345066427265633e-05,
+      "loss": 1.8853,
+      "step": 12450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7344659549397812e-05,
+      "loss": 1.9966,
+      "step": 12451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7344252645127677e-05,
+      "loss": 2.0646,
+      "step": 12452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.734384571445669e-05,
+      "loss": 1.7755,
+      "step": 12453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7343438757386315e-05,
+      "loss": 1.8551,
+      "step": 12454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7343031773918017e-05,
+      "loss": 1.7816,
+      "step": 12455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7342624764053254e-05,
+      "loss": 2.1692,
+      "step": 12456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7342217727793488e-05,
+      "loss": 2.0104,
+      "step": 12457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7341810665140192e-05,
+      "loss": 1.8937,
+      "step": 12458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.734140357609482e-05,
+      "loss": 1.7165,
+      "step": 12459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7340996460658836e-05,
+      "loss": 2.2559,
+      "step": 12460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7340589318833713e-05,
+      "loss": 1.725,
+      "step": 12461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7340182150620897e-05,
+      "loss": 2.2244,
+      "step": 12462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7339774956021868e-05,
+      "loss": 1.8865,
+      "step": 12463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7339367735038087e-05,
+      "loss": 2.0006,
+      "step": 12464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.733896048767101e-05,
+      "loss": 2.0886,
+      "step": 12465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7338553213922106e-05,
+      "loss": 1.6772,
+      "step": 12466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.733814591379284e-05,
+      "loss": 2.0297,
+      "step": 12467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7337738587284675e-05,
+      "loss": 1.8871,
+      "step": 12468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7337331234399074e-05,
+      "loss": 1.7523,
+      "step": 12469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7336923855137504e-05,
+      "loss": 1.7279,
+      "step": 12470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7336516449501428e-05,
+      "loss": 1.8038,
+      "step": 12471
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.7336109017492307e-05,
+      "loss": 2.213,
+      "step": 12472
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7335701559111612e-05,
+      "loss": 1.7839,
+      "step": 12473
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7335294074360806e-05,
+      "loss": 1.8037,
+      "step": 12474
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.733488656324135e-05,
+      "loss": 1.675,
+      "step": 12475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7334479025754712e-05,
+      "loss": 1.7843,
+      "step": 12476
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7334071461902355e-05,
+      "loss": 1.8909,
+      "step": 12477
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7333663871685748e-05,
+      "loss": 1.775,
+      "step": 12478
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7333256255106353e-05,
+      "loss": 2.0725,
+      "step": 12479
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7332848612165635e-05,
+      "loss": 1.9139,
+      "step": 12480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.7332440942865062e-05,
+      "loss": 2.2793,
+      "step": 12481
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7332033247206098e-05,
+      "loss": 2.1072,
+      "step": 12482
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7331625525190206e-05,
+      "loss": 2.3175,
+      "step": 12483
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7331217776818857e-05,
+      "loss": 1.8878,
+      "step": 12484
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.733081000209351e-05,
+      "loss": 1.5284,
+      "step": 12485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7330402201015638e-05,
+      "loss": 2.0077,
+      "step": 12486
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7329994373586702e-05,
+      "loss": 1.8339,
+      "step": 12487
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.732958651980817e-05,
+      "loss": 1.6556,
+      "step": 12488
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.732917863968151e-05,
+      "loss": 2.2507,
+      "step": 12489
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7328770733208186e-05,
+      "loss": 1.7217,
+      "step": 12490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7328362800389663e-05,
+      "loss": 1.6791,
+      "step": 12491
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.732795484122741e-05,
+      "loss": 2.0746,
+      "step": 12492
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.732754685572289e-05,
+      "loss": 1.9668,
+      "step": 12493
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7327138843877573e-05,
+      "loss": 1.7247,
+      "step": 12494
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7326730805692925e-05,
+      "loss": 2.0552,
+      "step": 12495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.7326322741170414e-05,
+      "loss": 2.017,
+      "step": 12496
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7325914650311505e-05,
+      "loss": 1.6779,
+      "step": 12497
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7325506533117663e-05,
+      "loss": 2.2301,
+      "step": 12498
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7325098389590357e-05,
+      "loss": 1.7768,
+      "step": 12499
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7324690219731056e-05,
+      "loss": 1.9728,
+      "step": 12500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7324282023541224e-05,
+      "loss": 2.0097,
+      "step": 12501
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7323873801022336e-05,
+      "loss": 1.6151,
+      "step": 12502
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.732346555217585e-05,
+      "loss": 2.0752,
+      "step": 12503
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7323057277003236e-05,
+      "loss": 1.7174,
+      "step": 12504
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7322648975505965e-05,
+      "loss": 2.0154,
+      "step": 12505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7322240647685503e-05,
+      "loss": 1.9151,
+      "step": 12506
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7321832293543315e-05,
+      "loss": 1.7824,
+      "step": 12507
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7321423913080876e-05,
+      "loss": 1.6713,
+      "step": 12508
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7321015506299646e-05,
+      "loss": 1.7805,
+      "step": 12509
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7320607073201096e-05,
+      "loss": 1.9987,
+      "step": 12510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7320198613786696e-05,
+      "loss": 1.8472,
+      "step": 12511
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7319790128057913e-05,
+      "loss": 1.6962,
+      "step": 12512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7319381616016214e-05,
+      "loss": 1.4892,
+      "step": 12513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7318973077663073e-05,
+      "loss": 1.7433,
+      "step": 12514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.731856451299995e-05,
+      "loss": 1.7183,
+      "step": 12515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7318155922028323e-05,
+      "loss": 1.948,
+      "step": 12516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.731774730474965e-05,
+      "loss": 1.6735,
+      "step": 12517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7317338661165416e-05,
+      "loss": 1.8287,
+      "step": 12518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.731692999127707e-05,
+      "loss": 1.6665,
+      "step": 12519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7316521295086096e-05,
+      "loss": 2.0067,
+      "step": 12520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.731611257259396e-05,
+      "loss": 1.6868,
+      "step": 12521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7315703823802126e-05,
+      "loss": 2.3261,
+      "step": 12522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7315295048712068e-05,
+      "loss": 1.6454,
+      "step": 12523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7314886247325253e-05,
+      "loss": 1.9845,
+      "step": 12524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7314477419643153e-05,
+      "loss": 1.6992,
+      "step": 12525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7314068565667237e-05,
+      "loss": 1.8322,
+      "step": 12526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.731365968539897e-05,
+      "loss": 1.8439,
+      "step": 12527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7313250778839835e-05,
+      "loss": 2.2029,
+      "step": 12528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7312841845991287e-05,
+      "loss": 1.7834,
+      "step": 12529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7312432886854803e-05,
+      "loss": 1.7137,
+      "step": 12530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7312023901431852e-05,
+      "loss": 1.6874,
+      "step": 12531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7311614889723903e-05,
+      "loss": 1.799,
+      "step": 12532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.375,
+      "learning_rate": 1.731120585173243e-05,
+      "loss": 1.9323,
+      "step": 12533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.7310796787458897e-05,
+      "loss": 1.7927,
+      "step": 12534
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.731038769690478e-05,
+      "loss": 1.7314,
+      "step": 12535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.7309978580071552e-05,
+      "loss": 1.6074,
+      "step": 12536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7309569436960675e-05,
+      "loss": 1.7478,
+      "step": 12537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.730916026757363e-05,
+      "loss": 1.7373,
+      "step": 12538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.730875107191188e-05,
+      "loss": 1.5059,
+      "step": 12539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7308341849976895e-05,
+      "loss": 1.8224,
+      "step": 12540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7307932601770154e-05,
+      "loss": 1.945,
+      "step": 12541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.730752332729312e-05,
+      "loss": 1.6225,
+      "step": 12542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.730711402654727e-05,
+      "loss": 1.8134,
+      "step": 12543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.730670469953407e-05,
+      "loss": 1.8605,
+      "step": 12544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7306295346255e-05,
+      "loss": 1.8173,
+      "step": 12545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7305885966711522e-05,
+      "loss": 1.9274,
+      "step": 12546
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7305476560905113e-05,
+      "loss": 1.8626,
+      "step": 12547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7305067128837243e-05,
+      "loss": 2.0123,
+      "step": 12548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7304657670509387e-05,
+      "loss": 1.9731,
+      "step": 12549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.730424818592301e-05,
+      "loss": 1.8682,
+      "step": 12550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.7303838675079593e-05,
+      "loss": 2.0838,
+      "step": 12551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.73034291379806e-05,
+      "loss": 1.7877,
+      "step": 12552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7303019574627503e-05,
+      "loss": 1.7014,
+      "step": 12553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7302609985021784e-05,
+      "loss": 1.7647,
+      "step": 12554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7302200369164906e-05,
+      "loss": 1.9579,
+      "step": 12555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7301790727058344e-05,
+      "loss": 1.4792,
+      "step": 12556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7301381058703572e-05,
+      "loss": 1.91,
+      "step": 12557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7300971364102062e-05,
+      "loss": 1.8243,
+      "step": 12558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7300561643255287e-05,
+      "loss": 2.0557,
+      "step": 12559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7300151896164717e-05,
+      "loss": 1.8037,
+      "step": 12560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.729974212283183e-05,
+      "loss": 1.6776,
+      "step": 12561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.7299332323258096e-05,
+      "loss": 2.1452,
+      "step": 12562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.729892249744499e-05,
+      "loss": 1.6633,
+      "step": 12563
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7298512645393983e-05,
+      "loss": 1.7115,
+      "step": 12564
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7298102767106547e-05,
+      "loss": 1.7975,
+      "step": 12565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7297692862584157e-05,
+      "loss": 1.5775,
+      "step": 12566
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7297282931828293e-05,
+      "loss": 2.0582,
+      "step": 12567
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7296872974840418e-05,
+      "loss": 1.6289,
+      "step": 12568
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.729646299162201e-05,
+      "loss": 2.0548,
+      "step": 12569
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7296052982174544e-05,
+      "loss": 2.0092,
+      "step": 12570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7295642946499493e-05,
+      "loss": 1.8345,
+      "step": 12571
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.729523288459833e-05,
+      "loss": 1.8917,
+      "step": 12572
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7294822796472534e-05,
+      "loss": 1.7404,
+      "step": 12573
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.729441268212357e-05,
+      "loss": 1.9906,
+      "step": 12574
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.729400254155292e-05,
+      "loss": 1.7092,
+      "step": 12575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7293592374762055e-05,
+      "loss": 1.9558,
+      "step": 12576
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.729318218175245e-05,
+      "loss": 2.0662,
+      "step": 12577
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.729277196252558e-05,
+      "loss": 2.1202,
+      "step": 12578
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7292361717082922e-05,
+      "loss": 1.6552,
+      "step": 12579
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.729195144542595e-05,
+      "loss": 1.7651,
+      "step": 12580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.729154114755613e-05,
+      "loss": 1.7279,
+      "step": 12581
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7291130823474945e-05,
+      "loss": 1.7785,
+      "step": 12582
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7290720473183875e-05,
+      "loss": 1.7446,
+      "step": 12583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7290310096684383e-05,
+      "loss": 1.5597,
+      "step": 12584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7289899693977955e-05,
+      "loss": 1.9007,
+      "step": 12585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7289489265066056e-05,
+      "loss": 2.3434,
+      "step": 12586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.7289078809950173e-05,
+      "loss": 1.9177,
+      "step": 12587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7288668328631774e-05,
+      "loss": 1.8675,
+      "step": 12588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.7288257821112334e-05,
+      "loss": 2.4013,
+      "step": 12589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.728784728739333e-05,
+      "loss": 1.7817,
+      "step": 12590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.728743672747624e-05,
+      "loss": 1.8673,
+      "step": 12591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7287026141362538e-05,
+      "loss": 2.3732,
+      "step": 12592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.72866155290537e-05,
+      "loss": 1.7681,
+      "step": 12593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7286204890551208e-05,
+      "loss": 1.5104,
+      "step": 12594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7285794225856528e-05,
+      "loss": 2.2273,
+      "step": 12595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7285383534971138e-05,
+      "loss": 1.5222,
+      "step": 12596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7284972817896523e-05,
+      "loss": 1.9383,
+      "step": 12597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.728456207463415e-05,
+      "loss": 1.9865,
+      "step": 12598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.72841513051855e-05,
+      "loss": 1.881,
+      "step": 12599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7283740509552047e-05,
+      "loss": 1.8824,
+      "step": 12600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.728332968773527e-05,
+      "loss": 1.8374,
+      "step": 12601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7282918839736645e-05,
+      "loss": 1.9728,
+      "step": 12602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7282507965557653e-05,
+      "loss": 1.8118,
+      "step": 12603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.728209706519976e-05,
+      "loss": 1.9465,
+      "step": 12604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7281686138664456e-05,
+      "loss": 2.0784,
+      "step": 12605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.7281275185953214e-05,
+      "loss": 1.9642,
+      "step": 12606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7280864207067505e-05,
+      "loss": 2.0177,
+      "step": 12607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7280453202008813e-05,
+      "loss": 1.9413,
+      "step": 12608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7280042170778612e-05,
+      "loss": 1.8225,
+      "step": 12609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.727963111337838e-05,
+      "loss": 1.9784,
+      "step": 12610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.72792200298096e-05,
+      "loss": 2.0789,
+      "step": 12611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.7278808920073743e-05,
+      "loss": 2.5184,
+      "step": 12612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.727839778417229e-05,
+      "loss": 1.7891,
+      "step": 12613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7277986622106717e-05,
+      "loss": 2.052,
+      "step": 12614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7277575433878506e-05,
+      "loss": 1.9253,
+      "step": 12615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.727716421948913e-05,
+      "loss": 1.9461,
+      "step": 12616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.727675297894007e-05,
+      "loss": 1.7908,
+      "step": 12617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7276341712232802e-05,
+      "loss": 1.7525,
+      "step": 12618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7275930419368808e-05,
+      "loss": 2.1613,
+      "step": 12619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7275519100349568e-05,
+      "loss": 2.1876,
+      "step": 12620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7275107755176557e-05,
+      "loss": 1.8768,
+      "step": 12621
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.727469638385125e-05,
+      "loss": 1.6616,
+      "step": 12622
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7274284986375133e-05,
+      "loss": 2.0431,
+      "step": 12623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.727387356274968e-05,
+      "loss": 1.9482,
+      "step": 12624
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7273462112976373e-05,
+      "loss": 1.6175,
+      "step": 12625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7273050637056687e-05,
+      "loss": 1.9576,
+      "step": 12626
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.727263913499211e-05,
+      "loss": 1.8159,
+      "step": 12627
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7272227606784108e-05,
+      "loss": 1.6574,
+      "step": 12628
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7271816052434172e-05,
+      "loss": 2.0023,
+      "step": 12629
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.727140447194378e-05,
+      "loss": 1.835,
+      "step": 12630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.72709928653144e-05,
+      "loss": 1.7611,
+      "step": 12631
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.7270581232547527e-05,
+      "loss": 2.0747,
+      "step": 12632
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7270169573644632e-05,
+      "loss": 2.003,
+      "step": 12633
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7269757888607196e-05,
+      "loss": 1.591,
+      "step": 12634
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7269346177436702e-05,
+      "loss": 2.0807,
+      "step": 12635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7268934440134623e-05,
+      "loss": 2.0625,
+      "step": 12636
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7268522676702447e-05,
+      "loss": 1.6799,
+      "step": 12637
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.726811088714165e-05,
+      "loss": 2.0669,
+      "step": 12638
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.7267699071453715e-05,
+      "loss": 1.9696,
+      "step": 12639
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.726728722964012e-05,
+      "loss": 1.9153,
+      "step": 12640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7266875361702348e-05,
+      "loss": 1.8617,
+      "step": 12641
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7266463467641873e-05,
+      "loss": 1.7724,
+      "step": 12642
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7266051547460182e-05,
+      "loss": 1.6237,
+      "step": 12643
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7265639601158754e-05,
+      "loss": 2.051,
+      "step": 12644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.726522762873907e-05,
+      "loss": 1.8754,
+      "step": 12645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7264815630202608e-05,
+      "loss": 1.9247,
+      "step": 12646
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7264403605550856e-05,
+      "loss": 2.2332,
+      "step": 12647
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.726399155478529e-05,
+      "loss": 1.8725,
+      "step": 12648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7263579477907394e-05,
+      "loss": 1.8116,
+      "step": 12649
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.7263167374918647e-05,
+      "loss": 1.8777,
+      "step": 12650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.726275524582053e-05,
+      "loss": 1.7596,
+      "step": 12651
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7262343090614524e-05,
+      "loss": 1.9463,
+      "step": 12652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7261930909302114e-05,
+      "loss": 2.3381,
+      "step": 12653
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7261518701884778e-05,
+      "loss": 2.0232,
+      "step": 12654
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7261106468364e-05,
+      "loss": 2.0293,
+      "step": 12655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7260694208741264e-05,
+      "loss": 1.9495,
+      "step": 12656
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7260281923018046e-05,
+      "loss": 1.8052,
+      "step": 12657
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7259869611195833e-05,
+      "loss": 1.8373,
+      "step": 12658
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7259457273276107e-05,
+      "loss": 1.8625,
+      "step": 12659
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.725904490926035e-05,
+      "loss": 1.9503,
+      "step": 12660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.725863251915004e-05,
+      "loss": 1.8403,
+      "step": 12661
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7258220102946665e-05,
+      "loss": 1.8295,
+      "step": 12662
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.72578076606517e-05,
+      "loss": 2.1431,
+      "step": 12663
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7257395192266638e-05,
+      "loss": 1.8913,
+      "step": 12664
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7256982697792957e-05,
+      "loss": 1.7096,
+      "step": 12665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.725657017723214e-05,
+      "loss": 2.05,
+      "step": 12666
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7256157630585666e-05,
+      "loss": 1.909,
+      "step": 12667
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.725574505785502e-05,
+      "loss": 2.0942,
+      "step": 12668
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.725533245904169e-05,
+      "loss": 1.7234,
+      "step": 12669
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7254919834147155e-05,
+      "loss": 1.8288,
+      "step": 12670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7254507183172896e-05,
+      "loss": 1.8907,
+      "step": 12671
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7254094506120406e-05,
+      "loss": 1.6282,
+      "step": 12672
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7253681802991156e-05,
+      "loss": 1.9629,
+      "step": 12673
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7253269073786637e-05,
+      "loss": 1.7221,
+      "step": 12674
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.725285631850833e-05,
+      "loss": 1.8996,
+      "step": 12675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7252443537157723e-05,
+      "loss": 1.7607,
+      "step": 12676
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7252030729736295e-05,
+      "loss": 1.8408,
+      "step": 12677
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.725161789624553e-05,
+      "loss": 2.4516,
+      "step": 12678
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7251205036686913e-05,
+      "loss": 1.6806,
+      "step": 12679
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7250792151061932e-05,
+      "loss": 2.1073,
+      "step": 12680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7250379239372066e-05,
+      "loss": 1.8472,
+      "step": 12681
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7249966301618803e-05,
+      "loss": 2.0106,
+      "step": 12682
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7249553337803624e-05,
+      "loss": 1.923,
+      "step": 12683
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7249140347928016e-05,
+      "loss": 1.7491,
+      "step": 12684
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7248727331993463e-05,
+      "loss": 1.7283,
+      "step": 12685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.724831429000145e-05,
+      "loss": 2.1262,
+      "step": 12686
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.7247901221953458e-05,
+      "loss": 2.1931,
+      "step": 12687
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7247488127850978e-05,
+      "loss": 2.1007,
+      "step": 12688
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7247075007695492e-05,
+      "loss": 1.9573,
+      "step": 12689
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7246661861488485e-05,
+      "loss": 2.2306,
+      "step": 12690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7246248689231443e-05,
+      "loss": 1.8444,
+      "step": 12691
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.724583549092585e-05,
+      "loss": 1.7522,
+      "step": 12692
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.724542226657319e-05,
+      "loss": 1.966,
+      "step": 12693
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7245009016174954e-05,
+      "loss": 1.8539,
+      "step": 12694
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.724459573973262e-05,
+      "loss": 1.8529,
+      "step": 12695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.724418243724768e-05,
+      "loss": 1.8084,
+      "step": 12696
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7243769108721616e-05,
+      "loss": 2.0,
+      "step": 12697
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.724335575415592e-05,
+      "loss": 1.896,
+      "step": 12698
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.724294237355207e-05,
+      "loss": 1.8333,
+      "step": 12699
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7242528966911555e-05,
+      "loss": 1.9409,
+      "step": 12700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.724211553423586e-05,
+      "loss": 1.9669,
+      "step": 12701
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7241702075526472e-05,
+      "loss": 1.9055,
+      "step": 12702
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.724128859078488e-05,
+      "loss": 1.8967,
+      "step": 12703
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7240875080012566e-05,
+      "loss": 1.8763,
+      "step": 12704
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.724046154321102e-05,
+      "loss": 1.6843,
+      "step": 12705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7240047980381727e-05,
+      "loss": 1.9491,
+      "step": 12706
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7239634391526173e-05,
+      "loss": 1.932,
+      "step": 12707
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7239220776645845e-05,
+      "loss": 1.8597,
+      "step": 12708
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.723880713574223e-05,
+      "loss": 1.8781,
+      "step": 12709
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7238393468816817e-05,
+      "loss": 1.7651,
+      "step": 12710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.723797977587109e-05,
+      "loss": 1.9566,
+      "step": 12711
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.723756605690654e-05,
+      "loss": 1.8353,
+      "step": 12712
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.723715231192465e-05,
+      "loss": 2.0882,
+      "step": 12713
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7236738540926907e-05,
+      "loss": 1.6892,
+      "step": 12714
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7236324743914802e-05,
+      "loss": 2.2124,
+      "step": 12715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7235910920889822e-05,
+      "loss": 1.8854,
+      "step": 12716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7235497071853452e-05,
+      "loss": 2.1963,
+      "step": 12717
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7235083196807184e-05,
+      "loss": 2.408,
+      "step": 12718
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.72346692957525e-05,
+      "loss": 1.6494,
+      "step": 12719
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.723425536869089e-05,
+      "loss": 2.0611,
+      "step": 12720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7233841415623844e-05,
+      "loss": 2.2551,
+      "step": 12721
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7233427436552847e-05,
+      "loss": 1.9123,
+      "step": 12722
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7233013431479395e-05,
+      "loss": 2.1417,
+      "step": 12723
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.7232599400404965e-05,
+      "loss": 1.9407,
+      "step": 12724
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7232185343331052e-05,
+      "loss": 2.0781,
+      "step": 12725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7231771260259144e-05,
+      "loss": 1.713,
+      "step": 12726
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7231357151190727e-05,
+      "loss": 1.7346,
+      "step": 12727
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7230943016127293e-05,
+      "loss": 1.8909,
+      "step": 12728
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7230528855070325e-05,
+      "loss": 1.8482,
+      "step": 12729
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7230114668021317e-05,
+      "loss": 2.1149,
+      "step": 12730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.722970045498176e-05,
+      "loss": 1.9001,
+      "step": 12731
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7229286215953137e-05,
+      "loss": 1.822,
+      "step": 12732
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7228871950936942e-05,
+      "loss": 2.1906,
+      "step": 12733
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.722845765993466e-05,
+      "loss": 1.8962,
+      "step": 12734
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7228043342947783e-05,
+      "loss": 1.8113,
+      "step": 12735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7227628999977798e-05,
+      "loss": 1.7508,
+      "step": 12736
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7227214631026198e-05,
+      "loss": 2.037,
+      "step": 12737
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7226800236094468e-05,
+      "loss": 2.1503,
+      "step": 12738
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7226385815184102e-05,
+      "loss": 1.8155,
+      "step": 12739
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.722597136829659e-05,
+      "loss": 1.9368,
+      "step": 12740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.722555689543342e-05,
+      "loss": 2.0079,
+      "step": 12741
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7225142396596076e-05,
+      "loss": 1.9977,
+      "step": 12742
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7224727871786057e-05,
+      "loss": 2.2826,
+      "step": 12743
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7224313321004853e-05,
+      "loss": 1.617,
+      "step": 12744
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7223898744253946e-05,
+      "loss": 1.7239,
+      "step": 12745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7223484141534833e-05,
+      "loss": 1.7195,
+      "step": 12746
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7223069512849006e-05,
+      "loss": 2.1131,
+      "step": 12747
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.722265485819795e-05,
+      "loss": 2.1889,
+      "step": 12748
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.722224017758316e-05,
+      "loss": 2.2,
+      "step": 12749
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.722182547100612e-05,
+      "loss": 1.9122,
+      "step": 12750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.722141073846833e-05,
+      "loss": 1.887,
+      "step": 12751
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.722099597997127e-05,
+      "loss": 2.2428,
+      "step": 12752
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7220581195516445e-05,
+      "loss": 1.9634,
+      "step": 12753
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7220166385105334e-05,
+      "loss": 2.0721,
+      "step": 12754
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.7219751548739433e-05,
+      "loss": 2.069,
+      "step": 12755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.721933668642023e-05,
+      "loss": 2.0256,
+      "step": 12756
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7218921798149222e-05,
+      "loss": 2.0445,
+      "step": 12757
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7218506883927898e-05,
+      "loss": 1.8425,
+      "step": 12758
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7218091943757746e-05,
+      "loss": 1.8414,
+      "step": 12759
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.721767697764026e-05,
+      "loss": 1.9891,
+      "step": 12760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.7217261985576938e-05,
+      "loss": 1.9674,
+      "step": 12761
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.721684696756926e-05,
+      "loss": 1.9069,
+      "step": 12762
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7216431923618727e-05,
+      "loss": 1.896,
+      "step": 12763
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7216016853726828e-05,
+      "loss": 2.1222,
+      "step": 12764
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7215601757895053e-05,
+      "loss": 1.9622,
+      "step": 12765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7215186636124897e-05,
+      "loss": 1.9772,
+      "step": 12766
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7214771488417853e-05,
+      "loss": 1.9435,
+      "step": 12767
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.721435631477541e-05,
+      "loss": 1.7032,
+      "step": 12768
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.721394111519906e-05,
+      "loss": 2.2113,
+      "step": 12769
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.72135258896903e-05,
+      "loss": 2.0754,
+      "step": 12770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7213110638250622e-05,
+      "loss": 2.4279,
+      "step": 12771
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7212695360881516e-05,
+      "loss": 1.9447,
+      "step": 12772
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7212280057584474e-05,
+      "loss": 1.8565,
+      "step": 12773
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.721186472836099e-05,
+      "loss": 1.7062,
+      "step": 12774
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7211449373212562e-05,
+      "loss": 1.9031,
+      "step": 12775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7211033992140677e-05,
+      "loss": 1.9206,
+      "step": 12776
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7210618585146833e-05,
+      "loss": 1.8984,
+      "step": 12777
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7210203152232517e-05,
+      "loss": 1.5758,
+      "step": 12778
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7209787693399226e-05,
+      "loss": 2.0917,
+      "step": 12779
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7209372208648455e-05,
+      "loss": 1.8037,
+      "step": 12780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7208956697981695e-05,
+      "loss": 1.5104,
+      "step": 12781
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.720854116140044e-05,
+      "loss": 2.3163,
+      "step": 12782
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7208125598906186e-05,
+      "loss": 1.7407,
+      "step": 12783
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.720771001050042e-05,
+      "loss": 1.8174,
+      "step": 12784
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7207294396184646e-05,
+      "loss": 2.3773,
+      "step": 12785
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7206878755960354e-05,
+      "loss": 1.7007,
+      "step": 12786
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7206463089829034e-05,
+      "loss": 1.8178,
+      "step": 12787
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7206047397792184e-05,
+      "loss": 1.5166,
+      "step": 12788
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7205631679851298e-05,
+      "loss": 1.7164,
+      "step": 12789
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.720521593600787e-05,
+      "loss": 1.8066,
+      "step": 12790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7204800166263398e-05,
+      "loss": 1.619,
+      "step": 12791
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.720438437061937e-05,
+      "loss": 1.9315,
+      "step": 12792
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7203968549077283e-05,
+      "loss": 1.8776,
+      "step": 12793
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7203552701638635e-05,
+      "loss": 1.9062,
+      "step": 12794
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.720313682830492e-05,
+      "loss": 1.8033,
+      "step": 12795
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.720272092907763e-05,
+      "loss": 2.1281,
+      "step": 12796
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7202305003958263e-05,
+      "loss": 1.7128,
+      "step": 12797
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.720188905294831e-05,
+      "loss": 2.2087,
+      "step": 12798
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7201473076049268e-05,
+      "loss": 1.9944,
+      "step": 12799
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7201057073262637e-05,
+      "loss": 1.7785,
+      "step": 12800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7200641044589906e-05,
+      "loss": 1.7923,
+      "step": 12801
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7200224990032577e-05,
+      "loss": 1.9356,
+      "step": 12802
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7199808909592137e-05,
+      "loss": 2.0097,
+      "step": 12803
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.719939280327009e-05,
+      "loss": 1.7998,
+      "step": 12804
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7198976671067928e-05,
+      "loss": 2.1398,
+      "step": 12805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7198560512987146e-05,
+      "loss": 1.9682,
+      "step": 12806
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.719814432902924e-05,
+      "loss": 1.821,
+      "step": 12807
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7197728119195713e-05,
+      "loss": 2.04,
+      "step": 12808
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7197311883488055e-05,
+      "loss": 1.9468,
+      "step": 12809
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7196895621907762e-05,
+      "loss": 2.1196,
+      "step": 12810
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7196479334456326e-05,
+      "loss": 1.6629,
+      "step": 12811
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7196063021135253e-05,
+      "loss": 1.799,
+      "step": 12812
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7195646681946034e-05,
+      "loss": 1.8798,
+      "step": 12813
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.719523031689017e-05,
+      "loss": 1.8372,
+      "step": 12814
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7194813925969153e-05,
+      "loss": 1.8789,
+      "step": 12815
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.719439750918448e-05,
+      "loss": 1.909,
+      "step": 12816
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.75,
+      "learning_rate": 1.719398106653765e-05,
+      "loss": 2.4886,
+      "step": 12817
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.719356459803016e-05,
+      "loss": 1.6817,
+      "step": 12818
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7193148103663505e-05,
+      "loss": 2.0319,
+      "step": 12819
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.7192731583439185e-05,
+      "loss": 2.0166,
+      "step": 12820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7192315037358696e-05,
+      "loss": 1.8229,
+      "step": 12821
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7191898465423532e-05,
+      "loss": 1.7668,
+      "step": 12822
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7191481867635202e-05,
+      "loss": 1.7488,
+      "step": 12823
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.719106524399519e-05,
+      "loss": 1.7855,
+      "step": 12824
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7190648594504998e-05,
+      "loss": 1.9359,
+      "step": 12825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7190231919166126e-05,
+      "loss": 1.6523,
+      "step": 12826
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.718981521798007e-05,
+      "loss": 2.0812,
+      "step": 12827
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.718939849094833e-05,
+      "loss": 1.8625,
+      "step": 12828
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.71889817380724e-05,
+      "loss": 1.9911,
+      "step": 12829
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7188564959353787e-05,
+      "loss": 1.7473,
+      "step": 12830
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7188148154793977e-05,
+      "loss": 2.1701,
+      "step": 12831
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7187731324394475e-05,
+      "loss": 1.9892,
+      "step": 12832
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7187314468156783e-05,
+      "loss": 1.9543,
+      "step": 12833
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7186897586082393e-05,
+      "loss": 1.9381,
+      "step": 12834
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7186480678172803e-05,
+      "loss": 1.6765,
+      "step": 12835
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.718606374442952e-05,
+      "loss": 1.9823,
+      "step": 12836
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7185646784854033e-05,
+      "loss": 1.864,
+      "step": 12837
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7185229799447848e-05,
+      "loss": 2.0156,
+      "step": 12838
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7184812788212457e-05,
+      "loss": 1.9158,
+      "step": 12839
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7184395751149367e-05,
+      "loss": 1.8251,
+      "step": 12840
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.718397868826007e-05,
+      "loss": 1.9494,
+      "step": 12841
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.718356159954607e-05,
+      "loss": 1.9319,
+      "step": 12842
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.7183144485008865e-05,
+      "loss": 1.9553,
+      "step": 12843
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7182727344649955e-05,
+      "loss": 1.8066,
+      "step": 12844
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.718231017847084e-05,
+      "loss": 1.9713,
+      "step": 12845
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7181892986473016e-05,
+      "loss": 1.8348,
+      "step": 12846
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.7181475768657988e-05,
+      "loss": 2.0507,
+      "step": 12847
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.718105852502725e-05,
+      "loss": 1.4548,
+      "step": 12848
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7180641255582307e-05,
+      "loss": 1.7628,
+      "step": 12849
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7180223960324656e-05,
+      "loss": 1.7742,
+      "step": 12850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.7179806639255797e-05,
+      "loss": 1.9342,
+      "step": 12851
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7179389292377234e-05,
+      "loss": 2.1172,
+      "step": 12852
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.717897191969046e-05,
+      "loss": 1.8056,
+      "step": 12853
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7178554521196983e-05,
+      "loss": 1.6148,
+      "step": 12854
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.71781370968983e-05,
+      "loss": 1.7425,
+      "step": 12855
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.717771964679591e-05,
+      "loss": 1.8225,
+      "step": 12856
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7177302170891317e-05,
+      "loss": 1.8946,
+      "step": 12857
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.717688466918602e-05,
+      "loss": 1.5711,
+      "step": 12858
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7176467141681518e-05,
+      "loss": 1.7659,
+      "step": 12859
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7176049588379314e-05,
+      "loss": 1.9288,
+      "step": 12860
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7175632009280912e-05,
+      "loss": 1.8464,
+      "step": 12861
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7175214404387806e-05,
+      "loss": 1.982,
+      "step": 12862
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7174796773701503e-05,
+      "loss": 1.9317,
+      "step": 12863
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.7174379117223504e-05,
+      "loss": 1.7026,
+      "step": 12864
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7173961434955307e-05,
+      "loss": 1.8412,
+      "step": 12865
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7173543726898415e-05,
+      "loss": 2.066,
+      "step": 12866
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.717312599305433e-05,
+      "loss": 2.2458,
+      "step": 12867
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.7172708233424555e-05,
+      "loss": 2.0992,
+      "step": 12868
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.717229044801059e-05,
+      "loss": 1.8137,
+      "step": 12869
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7171872636813933e-05,
+      "loss": 1.7098,
+      "step": 12870
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7171454799836093e-05,
+      "loss": 1.9282,
+      "step": 12871
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7171036937078567e-05,
+      "loss": 1.6882,
+      "step": 12872
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7170619048542865e-05,
+      "loss": 1.9862,
+      "step": 12873
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7170201134230476e-05,
+      "loss": 2.1316,
+      "step": 12874
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7169783194142916e-05,
+      "loss": 1.8169,
+      "step": 12875
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.7169365228281677e-05,
+      "loss": 2.1909,
+      "step": 12876
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.716894723664827e-05,
+      "loss": 1.7433,
+      "step": 12877
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.7168529219244185e-05,
+      "loss": 2.2574,
+      "step": 12878
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.716811117607094e-05,
+      "loss": 2.0353,
+      "step": 12879
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.716769310713003e-05,
+      "loss": 1.7264,
+      "step": 12880
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7167275012422954e-05,
+      "loss": 1.9785,
+      "step": 12881
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.716685689195122e-05,
+      "loss": 1.688,
+      "step": 12882
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.716643874571633e-05,
+      "loss": 1.6487,
+      "step": 12883
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7166020573719792e-05,
+      "loss": 1.6283,
+      "step": 12884
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.71656023759631e-05,
+      "loss": 1.8818,
+      "step": 12885
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7165184152447766e-05,
+      "loss": 1.9633,
+      "step": 12886
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.716476590317529e-05,
+      "loss": 1.6369,
+      "step": 12887
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.716434762814717e-05,
+      "loss": 2.0164,
+      "step": 12888
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.7163929327364918e-05,
+      "loss": 1.6075,
+      "step": 12889
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7163511000830034e-05,
+      "loss": 1.9174,
+      "step": 12890
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.7163092648544023e-05,
+      "loss": 1.8247,
+      "step": 12891
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7162674270508388e-05,
+      "loss": 1.6099,
+      "step": 12892
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.7162255866724627e-05,
+      "loss": 1.2421,
+      "step": 12893
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7161837437194257e-05,
+      "loss": 1.9514,
+      "step": 12894
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.7161418981918774e-05,
+      "loss": 2.2606,
+      "step": 12895
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.716100050089968e-05,
+      "loss": 1.7896,
+      "step": 12896
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7160581994138485e-05,
+      "loss": 1.8024,
+      "step": 12897
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.716016346163669e-05,
+      "loss": 1.9804,
+      "step": 12898
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7159744903395802e-05,
+      "loss": 1.8704,
+      "step": 12899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.7159326319417325e-05,
+      "loss": 1.7533,
+      "step": 12900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7158907709702762e-05,
+      "loss": 1.7926,
+      "step": 12901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.715848907425362e-05,
+      "loss": 1.8509,
+      "step": 12902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.71580704130714e-05,
+      "loss": 2.027,
+      "step": 12903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.715765172615761e-05,
+      "loss": 1.8546,
+      "step": 12904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7157233013513756e-05,
+      "loss": 1.7758,
+      "step": 12905
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.7156814275141346e-05,
+      "loss": 1.8971,
+      "step": 12906
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7156395511041876e-05,
+      "loss": 1.9513,
+      "step": 12907
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.715597672121686e-05,
+      "loss": 2.0471,
+      "step": 12908
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7155557905667795e-05,
+      "loss": 2.0047,
+      "step": 12909
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7155139064396196e-05,
+      "loss": 1.6564,
+      "step": 12910
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.715472019740356e-05,
+      "loss": 1.8438,
+      "step": 12911
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.71543013046914e-05,
+      "loss": 1.9088,
+      "step": 12912
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.715388238626122e-05,
+      "loss": 1.7122,
+      "step": 12913
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7153463442114522e-05,
+      "loss": 2.1035,
+      "step": 12914
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7153044472252815e-05,
+      "loss": 1.9152,
+      "step": 12915
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7152625476677607e-05,
+      "loss": 1.917,
+      "step": 12916
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.71522064553904e-05,
+      "loss": 2.0858,
+      "step": 12917
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.7151787408392702e-05,
+      "loss": 1.7371,
+      "step": 12918
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.715136833568602e-05,
+      "loss": 1.8081,
+      "step": 12919
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.715094923727186e-05,
+      "loss": 1.6236,
+      "step": 12920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.715053011315173e-05,
+      "loss": 2.0189,
+      "step": 12921
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7150110963327133e-05,
+      "loss": 2.0397,
+      "step": 12922
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.714969178779958e-05,
+      "loss": 1.7712,
+      "step": 12923
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7149272586570576e-05,
+      "loss": 1.8267,
+      "step": 12924
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7148853359641627e-05,
+      "loss": 2.1295,
+      "step": 12925
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.714843410701424e-05,
+      "loss": 1.5861,
+      "step": 12926
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7148014828689924e-05,
+      "loss": 1.9078,
+      "step": 12927
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7147595524670183e-05,
+      "loss": 1.6935,
+      "step": 12928
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.714717619495653e-05,
+      "loss": 1.6404,
+      "step": 12929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7146756839550468e-05,
+      "loss": 1.9352,
+      "step": 12930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7146337458453503e-05,
+      "loss": 1.965,
+      "step": 12931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7145918051667147e-05,
+      "loss": 2.0329,
+      "step": 12932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7145498619192907e-05,
+      "loss": 2.2195,
+      "step": 12933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7145079161032285e-05,
+      "loss": 1.8578,
+      "step": 12934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.71446596771868e-05,
+      "loss": 1.5136,
+      "step": 12935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7144240167657945e-05,
+      "loss": 1.7054,
+      "step": 12936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.7143820632447238e-05,
+      "loss": 2.3671,
+      "step": 12937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7143401071556187e-05,
+      "loss": 2.1698,
+      "step": 12938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7142981484986298e-05,
+      "loss": 1.8016,
+      "step": 12939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7142561872739083e-05,
+      "loss": 2.0209,
+      "step": 12940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7142142234816044e-05,
+      "loss": 2.0875,
+      "step": 12941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7141722571218693e-05,
+      "loss": 2.0194,
+      "step": 12942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.714130288194854e-05,
+      "loss": 1.7818,
+      "step": 12943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7140883167007088e-05,
+      "loss": 1.8217,
+      "step": 12944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7140463426395855e-05,
+      "loss": 1.731,
+      "step": 12945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.714004366011634e-05,
+      "loss": 2.4995,
+      "step": 12946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7139623868170062e-05,
+      "loss": 1.6746,
+      "step": 12947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.713920405055852e-05,
+      "loss": 2.2382,
+      "step": 12948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.713878420728323e-05,
+      "loss": 2.6216,
+      "step": 12949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7138364338345698e-05,
+      "loss": 1.8069,
+      "step": 12950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7137944443747433e-05,
+      "loss": 1.8728,
+      "step": 12951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.713752452348995e-05,
+      "loss": 1.9768,
+      "step": 12952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.713710457757475e-05,
+      "loss": 1.8679,
+      "step": 12953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.7136684606003347e-05,
+      "loss": 1.9422,
+      "step": 12954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7136264608777252e-05,
+      "loss": 1.7767,
+      "step": 12955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7135844585897975e-05,
+      "loss": 1.8813,
+      "step": 12956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7135424537367023e-05,
+      "loss": 2.1783,
+      "step": 12957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.713500446318591e-05,
+      "loss": 1.8627,
+      "step": 12958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7134584363356138e-05,
+      "loss": 1.7869,
+      "step": 12959
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.7134164237879226e-05,
+      "loss": 1.7548,
+      "step": 12960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.713374408675668e-05,
+      "loss": 1.9519,
+      "step": 12961
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.713332390999001e-05,
+      "loss": 1.4966,
+      "step": 12962
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.713290370758073e-05,
+      "loss": 2.1291,
+      "step": 12963
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7132483479530345e-05,
+      "loss": 1.7763,
+      "step": 12964
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.713206322584037e-05,
+      "loss": 1.8194,
+      "step": 12965
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7131642946512314e-05,
+      "loss": 1.7419,
+      "step": 12966
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.713122264154769e-05,
+      "loss": 1.8437,
+      "step": 12967
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7130802310948005e-05,
+      "loss": 1.8298,
+      "step": 12968
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7130381954714775e-05,
+      "loss": 1.937,
+      "step": 12969
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7129961572849504e-05,
+      "loss": 2.1089,
+      "step": 12970
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.712954116535371e-05,
+      "loss": 1.9699,
+      "step": 12971
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.71291207322289e-05,
+      "loss": 2.2384,
+      "step": 12972
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7128700273476592e-05,
+      "loss": 1.9204,
+      "step": 12973
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.7128279789098285e-05,
+      "loss": 1.7257,
+      "step": 12974
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7127859279095503e-05,
+      "loss": 1.8164,
+      "step": 12975
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.712743874346975e-05,
+      "loss": 2.2899,
+      "step": 12976
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.712701818222254e-05,
+      "loss": 1.717,
+      "step": 12977
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.712659759535539e-05,
+      "loss": 2.3615,
+      "step": 12978
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.71261769828698e-05,
+      "loss": 1.8187,
+      "step": 12979
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7125756344767293e-05,
+      "loss": 1.8898,
+      "step": 12980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7125335681049376e-05,
+      "loss": 1.9362,
+      "step": 12981
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7124914991717566e-05,
+      "loss": 1.9364,
+      "step": 12982
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7124494276773367e-05,
+      "loss": 2.1108,
+      "step": 12983
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7124073536218297e-05,
+      "loss": 1.6285,
+      "step": 12984
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.712365277005387e-05,
+      "loss": 1.9033,
+      "step": 12985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.712323197828159e-05,
+      "loss": 2.4451,
+      "step": 12986
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.712281116090298e-05,
+      "loss": 1.7918,
+      "step": 12987
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.712239031791955e-05,
+      "loss": 1.7324,
+      "step": 12988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7121969449332806e-05,
+      "loss": 1.766,
+      "step": 12989
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7121548555144272e-05,
+      "loss": 1.8426,
+      "step": 12990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.712112763535545e-05,
+      "loss": 2.1358,
+      "step": 12991
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.712070668996786e-05,
+      "loss": 2.0485,
+      "step": 12992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7120285718983018e-05,
+      "loss": 1.6191,
+      "step": 12993
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7119864722402427e-05,
+      "loss": 1.925,
+      "step": 12994
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.711944370022761e-05,
+      "loss": 2.0449,
+      "step": 12995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7119022652460074e-05,
+      "loss": 1.9871,
+      "step": 12996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7118601579101334e-05,
+      "loss": 2.0424,
+      "step": 12997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7118180480152905e-05,
+      "loss": 1.9795,
+      "step": 12998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7117759355616302e-05,
+      "loss": 1.6735,
+      "step": 12999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.711733820549304e-05,
+      "loss": 1.9445,
+      "step": 13000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7116917029784628e-05,
+      "loss": 1.6702,
+      "step": 13001
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7116495828492583e-05,
+      "loss": 2.0598,
+      "step": 13002
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.7116074601618418e-05,
+      "loss": 1.964,
+      "step": 13003
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7115653349163645e-05,
+      "loss": 1.9898,
+      "step": 13004
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7115232071129787e-05,
+      "loss": 1.964,
+      "step": 13005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7114810767518348e-05,
+      "loss": 2.1437,
+      "step": 13006
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7114389438330846e-05,
+      "loss": 1.7503,
+      "step": 13007
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.71139680835688e-05,
+      "loss": 2.0113,
+      "step": 13008
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7113546703233716e-05,
+      "loss": 2.0904,
+      "step": 13009
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.711312529732712e-05,
+      "loss": 1.974,
+      "step": 13010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7112703865850517e-05,
+      "loss": 1.7268,
+      "step": 13011
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0,
+      "learning_rate": 1.711228240880543e-05,
+      "loss": 2.2841,
+      "step": 13012
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7111860926193365e-05,
+      "loss": 2.0497,
+      "step": 13013
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7111439418015842e-05,
+      "loss": 1.8551,
+      "step": 13014
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.711101788427438e-05,
+      "loss": 1.9518,
+      "step": 13015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7110596324970486e-05,
+      "loss": 1.7806,
+      "step": 13016
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7110174740105683e-05,
+      "loss": 1.8435,
+      "step": 13017
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.710975312968148e-05,
+      "loss": 1.7524,
+      "step": 13018
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7109331493699398e-05,
+      "loss": 2.1355,
+      "step": 13019
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7108909832160952e-05,
+      "loss": 2.2589,
+      "step": 13020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7108488145067654e-05,
+      "loss": 1.9913,
+      "step": 13021
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7108066432421023e-05,
+      "loss": 1.6683,
+      "step": 13022
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7107644694222575e-05,
+      "loss": 1.628,
+      "step": 13023
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7107222930473825e-05,
+      "loss": 1.5922,
+      "step": 13024
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.710680114117629e-05,
+      "loss": 1.8878,
+      "step": 13025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7106379326331483e-05,
+      "loss": 1.7814,
+      "step": 13026
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.7105957485940926e-05,
+      "loss": 1.4875,
+      "step": 13027
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7105535620006128e-05,
+      "loss": 1.8675,
+      "step": 13028
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7105113728528615e-05,
+      "loss": 2.1586,
+      "step": 13029
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7104691811509894e-05,
+      "loss": 2.1139,
+      "step": 13030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.710426986895149e-05,
+      "loss": 1.4513,
+      "step": 13031
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.7103847900854914e-05,
+      "loss": 1.701,
+      "step": 13032
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7103425907221683e-05,
+      "loss": 2.1039,
+      "step": 13033
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.710300388805332e-05,
+      "loss": 2.2752,
+      "step": 13034
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7102581843351335e-05,
+      "loss": 1.9155,
+      "step": 13035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.710215977311725e-05,
+      "loss": 1.779,
+      "step": 13036
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.710173767735258e-05,
+      "loss": 1.7065,
+      "step": 13037
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7101315556058837e-05,
+      "loss": 1.8505,
+      "step": 13038
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.710089340923755e-05,
+      "loss": 1.7823,
+      "step": 13039
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.710047123689023e-05,
+      "loss": 1.5879,
+      "step": 13040
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7100049039018393e-05,
+      "loss": 1.8318,
+      "step": 13041
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.709962681562356e-05,
+      "loss": 1.6615,
+      "step": 13042
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.709920456670725e-05,
+      "loss": 1.7433,
+      "step": 13043
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7098782292270972e-05,
+      "loss": 2.1688,
+      "step": 13044
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7098359992316255e-05,
+      "loss": 1.5911,
+      "step": 13045
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7097937666844613e-05,
+      "loss": 2.3755,
+      "step": 13046
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7097515315857564e-05,
+      "loss": 1.894,
+      "step": 13047
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.7097092939356622e-05,
+      "loss": 1.8572,
+      "step": 13048
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.7096670537343315e-05,
+      "loss": 2.1554,
+      "step": 13049
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7096248109819152e-05,
+      "loss": 2.2755,
+      "step": 13050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.7095825656785656e-05,
+      "loss": 1.9584,
+      "step": 13051
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7095403178244345e-05,
+      "loss": 1.942,
+      "step": 13052
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7094980674196735e-05,
+      "loss": 1.5953,
+      "step": 13053
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7094558144644353e-05,
+      "loss": 1.9422,
+      "step": 13054
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.709413558958871e-05,
+      "loss": 1.9783,
+      "step": 13055
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.7093713009031324e-05,
+      "loss": 2.1559,
+      "step": 13056
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7093290402973723e-05,
+      "loss": 1.81,
+      "step": 13057
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.7092867771417415e-05,
+      "loss": 1.8426,
+      "step": 13058
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.7092445114363924e-05,
+      "loss": 2.1516,
+      "step": 13059
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7092022431814776e-05,
+      "loss": 2.088,
+      "step": 13060
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.709159972377148e-05,
+      "loss": 2.3734,
+      "step": 13061
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.7091176990235564e-05,
+      "loss": 1.6204,
+      "step": 13062
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.709075423120854e-05,
+      "loss": 2.088,
+      "step": 13063
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.7090331446691935e-05,
+      "loss": 1.8189,
+      "step": 13064
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.708990863668726e-05,
+      "loss": 1.8956,
+      "step": 13065
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7089485801196042e-05,
+      "loss": 1.8309,
+      "step": 13066
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.7089062940219804e-05,
+      "loss": 1.8442,
+      "step": 13067
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7088640053760057e-05,
+      "loss": 1.892,
+      "step": 13068
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.708821714181833e-05,
+      "loss": 1.9932,
+      "step": 13069
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.708779420439613e-05,
+      "loss": 1.841,
+      "step": 13070
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.708737124149499e-05,
+      "loss": 1.8704,
+      "step": 13071
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7086948253116432e-05,
+      "loss": 1.7129,
+      "step": 13072
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.7086525239261964e-05,
+      "loss": 2.1426,
+      "step": 13073
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7086102199933116e-05,
+      "loss": 1.8164,
+      "step": 13074
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.708567913513141e-05,
+      "loss": 1.9415,
+      "step": 13075
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7085256044858357e-05,
+      "loss": 1.9712,
+      "step": 13076
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.708483292911549e-05,
+      "loss": 1.9337,
+      "step": 13077
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7084409787904318e-05,
+      "loss": 1.7979,
+      "step": 13078
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.708398662122637e-05,
+      "loss": 1.9618,
+      "step": 13079
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7083563429083166e-05,
+      "loss": 1.9466,
+      "step": 13080
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.708314021147623e-05,
+      "loss": 2.0279,
+      "step": 13081
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.708271696840708e-05,
+      "loss": 1.9583,
+      "step": 13082
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7082293699877234e-05,
+      "loss": 1.7876,
+      "step": 13083
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7081870405888218e-05,
+      "loss": 2.0145,
+      "step": 13084
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.7081447086441553e-05,
+      "loss": 1.9926,
+      "step": 13085
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.708102374153876e-05,
+      "loss": 1.8963,
+      "step": 13086
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.708060037118136e-05,
+      "loss": 1.7472,
+      "step": 13087
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7080176975370877e-05,
+      "loss": 1.9847,
+      "step": 13088
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7079753554108832e-05,
+      "loss": 1.9375,
+      "step": 13089
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.707933010739675e-05,
+      "loss": 1.7138,
+      "step": 13090
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7078906635236147e-05,
+      "loss": 1.9784,
+      "step": 13091
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7078483137628552e-05,
+      "loss": 1.9533,
+      "step": 13092
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7078059614575485e-05,
+      "loss": 1.8218,
+      "step": 13093
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7077636066078462e-05,
+      "loss": 1.9549,
+      "step": 13094
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7077212492139013e-05,
+      "loss": 2.2916,
+      "step": 13095
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.707678889275866e-05,
+      "loss": 2.0912,
+      "step": 13096
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.7076365267938926e-05,
+      "loss": 1.8322,
+      "step": 13097
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.707594161768133e-05,
+      "loss": 2.0884,
+      "step": 13098
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.70755179419874e-05,
+      "loss": 1.6435,
+      "step": 13099
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7075094240858656e-05,
+      "loss": 1.9543,
+      "step": 13100
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.707467051429662e-05,
+      "loss": 2.0088,
+      "step": 13101
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.7074246762302814e-05,
+      "loss": 1.5868,
+      "step": 13102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7073822984878766e-05,
+      "loss": 1.9455,
+      "step": 13103
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.7073399182026e-05,
+      "loss": 1.844,
+      "step": 13104
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7072975353746037e-05,
+      "loss": 1.7247,
+      "step": 13105
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7072551500040394e-05,
+      "loss": 1.9009,
+      "step": 13106
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.7072127620910606e-05,
+      "loss": 1.8721,
+      "step": 13107
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.707170371635819e-05,
+      "loss": 1.9992,
+      "step": 13108
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7071279786384673e-05,
+      "loss": 1.9268,
+      "step": 13109
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7070855830991576e-05,
+      "loss": 2.1878,
+      "step": 13110
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7070431850180423e-05,
+      "loss": 1.8245,
+      "step": 13111
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7070007843952746e-05,
+      "loss": 1.7775,
+      "step": 13112
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.7069583812310058e-05,
+      "loss": 1.9326,
+      "step": 13113
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7069159755253885e-05,
+      "loss": 1.7715,
+      "step": 13114
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.706873567278576e-05,
+      "loss": 1.8406,
+      "step": 13115
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.70683115649072e-05,
+      "loss": 1.7793,
+      "step": 13116
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.706788743161973e-05,
+      "loss": 2.2919,
+      "step": 13117
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7067463272924882e-05,
+      "loss": 1.8208,
+      "step": 13118
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.706703908882417e-05,
+      "loss": 1.7878,
+      "step": 13119
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7066614879319124e-05,
+      "loss": 2.1766,
+      "step": 13120
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7066190644411267e-05,
+      "loss": 1.6719,
+      "step": 13121
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.706576638410213e-05,
+      "loss": 1.8377,
+      "step": 13122
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.7065342098393233e-05,
+      "loss": 1.7572,
+      "step": 13123
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.70649177872861e-05,
+      "loss": 1.7404,
+      "step": 13124
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7064493450782257e-05,
+      "loss": 2.0282,
+      "step": 13125
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7064069088883236e-05,
+      "loss": 1.9113,
+      "step": 13126
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.706364470159055e-05,
+      "loss": 1.7593,
+      "step": 13127
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.706322028890574e-05,
+      "loss": 1.8877,
+      "step": 13128
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.706279585083032e-05,
+      "loss": 2.1824,
+      "step": 13129
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7062371387365817e-05,
+      "loss": 1.9884,
+      "step": 13130
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7061946898513762e-05,
+      "loss": 1.894,
+      "step": 13131
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7061522384275676e-05,
+      "loss": 1.7541,
+      "step": 13132
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.706109784465309e-05,
+      "loss": 2.4333,
+      "step": 13133
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7060673279647525e-05,
+      "loss": 1.7462,
+      "step": 13134
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.706024868926051e-05,
+      "loss": 1.8123,
+      "step": 13135
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7059824073493567e-05,
+      "loss": 1.9063,
+      "step": 13136
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.705939943234823e-05,
+      "loss": 2.1166,
+      "step": 13137
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.705897476582602e-05,
+      "loss": 1.7009,
+      "step": 13138
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.7058550073928466e-05,
+      "loss": 1.9417,
+      "step": 13139
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.705812535665709e-05,
+      "loss": 1.9837,
+      "step": 13140
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.705770061401343e-05,
+      "loss": 1.8083,
+      "step": 13141
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7057275845999003e-05,
+      "loss": 1.8867,
+      "step": 13142
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7056851052615334e-05,
+      "loss": 1.743,
+      "step": 13143
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7056426233863956e-05,
+      "loss": 1.7019,
+      "step": 13144
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.70560013897464e-05,
+      "loss": 2.0206,
+      "step": 13145
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.705557652026418e-05,
+      "loss": 2.046,
+      "step": 13146
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7055151625418833e-05,
+      "loss": 1.7924,
+      "step": 13147
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.705472670521189e-05,
+      "loss": 1.5672,
+      "step": 13148
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.7054301759644866e-05,
+      "loss": 2.3338,
+      "step": 13149
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.70538767887193e-05,
+      "loss": 1.8302,
+      "step": 13150
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.705345179243671e-05,
+      "loss": 1.8895,
+      "step": 13151
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7053026770798636e-05,
+      "loss": 1.7726,
+      "step": 13152
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7052601723806593e-05,
+      "loss": 2.0583,
+      "step": 13153
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7052176651462117e-05,
+      "loss": 2.1931,
+      "step": 13154
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7051751553766733e-05,
+      "loss": 1.9196,
+      "step": 13155
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.7051326430721968e-05,
+      "loss": 2.8783,
+      "step": 13156
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.7050901282329355e-05,
+      "loss": 2.0011,
+      "step": 13157
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.705047610859042e-05,
+      "loss": 1.8512,
+      "step": 13158
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7050050909506687e-05,
+      "loss": 1.6563,
+      "step": 13159
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7049625685079694e-05,
+      "loss": 1.9275,
+      "step": 13160
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.7049200435310956e-05,
+      "loss": 2.0941,
+      "step": 13161
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7048775160202018e-05,
+      "loss": 2.2159,
+      "step": 13162
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7048349859754398e-05,
+      "loss": 1.6816,
+      "step": 13163
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7047924533969623e-05,
+      "loss": 1.6229,
+      "step": 13164
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.704749918284923e-05,
+      "loss": 1.8756,
+      "step": 13165
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7047073806394743e-05,
+      "loss": 1.9932,
+      "step": 13166
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7046648404607688e-05,
+      "loss": 1.9951,
+      "step": 13167
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7046222977489604e-05,
+      "loss": 1.4733,
+      "step": 13168
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.7045797525042013e-05,
+      "loss": 2.1406,
+      "step": 13169
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.704537204726645e-05,
+      "loss": 1.6512,
+      "step": 13170
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7044946544164434e-05,
+      "loss": 1.7834,
+      "step": 13171
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7044521015737506e-05,
+      "loss": 1.5912,
+      "step": 13172
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.704409546198719e-05,
+      "loss": 2.018,
+      "step": 13173
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.7043669882915014e-05,
+      "loss": 1.5482,
+      "step": 13174
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7043244278522512e-05,
+      "loss": 1.5787,
+      "step": 13175
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7042818648811213e-05,
+      "loss": 1.575,
+      "step": 13176
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7042392993782645e-05,
+      "loss": 2.0124,
+      "step": 13177
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.7041967313438342e-05,
+      "loss": 2.0247,
+      "step": 13178
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.7041541607779833e-05,
+      "loss": 2.2413,
+      "step": 13179
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7041115876808645e-05,
+      "loss": 1.846,
+      "step": 13180
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.704069012052631e-05,
+      "loss": 1.7843,
+      "step": 13181
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.704026433893436e-05,
+      "loss": 2.0697,
+      "step": 13182
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7039838532034326e-05,
+      "loss": 1.6304,
+      "step": 13183
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7039412699827736e-05,
+      "loss": 1.8766,
+      "step": 13184
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.703898684231612e-05,
+      "loss": 1.7054,
+      "step": 13185
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7038560959501013e-05,
+      "loss": 2.0776,
+      "step": 13186
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.7038135051383947e-05,
+      "loss": 1.7594,
+      "step": 13187
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.7037709117966447e-05,
+      "loss": 1.7374,
+      "step": 13188
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7037283159250045e-05,
+      "loss": 2.037,
+      "step": 13189
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7036857175236275e-05,
+      "loss": 1.9018,
+      "step": 13190
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7036431165926666e-05,
+      "loss": 1.835,
+      "step": 13191
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7036005131322756e-05,
+      "loss": 1.9788,
+      "step": 13192
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7035579071426073e-05,
+      "loss": 1.7709,
+      "step": 13193
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.703515298623814e-05,
+      "loss": 1.3521,
+      "step": 13194
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7034726875760503e-05,
+      "loss": 2.0016,
+      "step": 13195
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7034300739994682e-05,
+      "loss": 2.3682,
+      "step": 13196
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7033874578942217e-05,
+      "loss": 2.3578,
+      "step": 13197
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.703344839260463e-05,
+      "loss": 2.0112,
+      "step": 13198
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7033022180983464e-05,
+      "loss": 1.5815,
+      "step": 13199
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7032595944080246e-05,
+      "loss": 1.5591,
+      "step": 13200
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7032169681896506e-05,
+      "loss": 2.0674,
+      "step": 13201
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7031743394433783e-05,
+      "loss": 1.7483,
+      "step": 13202
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.7031317081693604e-05,
+      "loss": 2.182,
+      "step": 13203
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.70308907436775e-05,
+      "loss": 1.6462,
+      "step": 13204
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.703046438038701e-05,
+      "loss": 1.9479,
+      "step": 13205
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7030037991823664e-05,
+      "loss": 1.7372,
+      "step": 13206
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.7029611577988995e-05,
+      "loss": 2.4875,
+      "step": 13207
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.702918513888453e-05,
+      "loss": 1.8936,
+      "step": 13208
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.702875867451181e-05,
+      "loss": 1.8827,
+      "step": 13209
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7028332184872366e-05,
+      "loss": 2.2938,
+      "step": 13210
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.7027905669967727e-05,
+      "loss": 1.8936,
+      "step": 13211
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.702747912979943e-05,
+      "loss": 1.6477,
+      "step": 13212
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7027052564369007e-05,
+      "loss": 1.7674,
+      "step": 13213
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.702662597367799e-05,
+      "loss": 1.6842,
+      "step": 13214
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.702619935772792e-05,
+      "loss": 1.9961,
+      "step": 13215
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.7025772716520324e-05,
+      "loss": 1.8451,
+      "step": 13216
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.7025346050056734e-05,
+      "loss": 2.0034,
+      "step": 13217
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7024919358338687e-05,
+      "loss": 1.831,
+      "step": 13218
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.7024492641367714e-05,
+      "loss": 2.1488,
+      "step": 13219
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.702406589914535e-05,
+      "loss": 1.9565,
+      "step": 13220
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7023639131673136e-05,
+      "loss": 1.4957,
+      "step": 13221
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7023212338952597e-05,
+      "loss": 2.1113,
+      "step": 13222
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.702278552098527e-05,
+      "loss": 1.9525,
+      "step": 13223
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7022358677772695e-05,
+      "loss": 1.7946,
+      "step": 13224
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.7021931809316395e-05,
+      "loss": 1.5718,
+      "step": 13225
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.375,
+      "learning_rate": 1.7021504915617914e-05,
+      "loss": 1.9388,
+      "step": 13226
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7021077996678786e-05,
+      "loss": 1.9114,
+      "step": 13227
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.702065105250054e-05,
+      "loss": 1.9421,
+      "step": 13228
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.7020224083084712e-05,
+      "loss": 1.7909,
+      "step": 13229
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.701979708843284e-05,
+      "loss": 2.0244,
+      "step": 13230
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.375,
+      "learning_rate": 1.701937006854646e-05,
+      "loss": 2.0296,
+      "step": 13231
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.70189430234271e-05,
+      "loss": 2.2362,
+      "step": 13232
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.7018515953076303e-05,
+      "loss": 1.9909,
+      "step": 13233
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.70180888574956e-05,
+      "loss": 2.0584,
+      "step": 13234
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.7017661736686528e-05,
+      "loss": 1.8591,
+      "step": 13235
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.701723459065062e-05,
+      "loss": 1.7584,
+      "step": 13236
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7016807419389416e-05,
+      "loss": 1.7011,
+      "step": 13237
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7016380222904446e-05,
+      "loss": 1.5627,
+      "step": 13238
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.701595300119725e-05,
+      "loss": 1.9106,
+      "step": 13239
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7015525754269363e-05,
+      "loss": 1.8908,
+      "step": 13240
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.7015098482122318e-05,
+      "loss": 2.0083,
+      "step": 13241
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7014671184757654e-05,
+      "loss": 2.0029,
+      "step": 13242
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.701424386217691e-05,
+      "loss": 1.6556,
+      "step": 13243
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.7013816514381608e-05,
+      "loss": 1.7545,
+      "step": 13244
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.70133891413733e-05,
+      "loss": 1.8597,
+      "step": 13245
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.701296174315352e-05,
+      "loss": 1.888,
+      "step": 13246
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7012534319723797e-05,
+      "loss": 1.7857,
+      "step": 13247
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.7012106871085673e-05,
+      "loss": 1.7095,
+      "step": 13248
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.7011679397240683e-05,
+      "loss": 1.7583,
+      "step": 13249
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.7011251898190363e-05,
+      "loss": 1.8967,
+      "step": 13250
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.701082437393625e-05,
+      "loss": 2.1052,
+      "step": 13251
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7010396824479883e-05,
+      "loss": 1.8895,
+      "step": 13252
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7009969249822796e-05,
+      "loss": 1.9755,
+      "step": 13253
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7009541649966527e-05,
+      "loss": 2.0545,
+      "step": 13254
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.7009114024912616e-05,
+      "loss": 2.1679,
+      "step": 13255
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.70086863746626e-05,
+      "loss": 1.9672,
+      "step": 13256
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.7008258699218006e-05,
+      "loss": 1.7065,
+      "step": 13257
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.7007830998580384e-05,
+      "loss": 1.8356,
+      "step": 13258
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.7007403272751267e-05,
+      "loss": 2.0327,
+      "step": 13259
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.7006975521732193e-05,
+      "loss": 1.8861,
+      "step": 13260
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7006547745524698e-05,
+      "loss": 1.9311,
+      "step": 13261
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7006119944130318e-05,
+      "loss": 1.9079,
+      "step": 13262
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.70056921175506e-05,
+      "loss": 1.9878,
+      "step": 13263
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7005264265787073e-05,
+      "loss": 1.8203,
+      "step": 13264
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7004836388841274e-05,
+      "loss": 2.0347,
+      "step": 13265
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7004408486714748e-05,
+      "loss": 2.1734,
+      "step": 13266
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.700398055940903e-05,
+      "loss": 1.7012,
+      "step": 13267
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.700355260692566e-05,
+      "loss": 1.8906,
+      "step": 13268
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.700312462926617e-05,
+      "loss": 2.1228,
+      "step": 13269
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7002696626432108e-05,
+      "loss": 1.7835,
+      "step": 13270
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.7002268598425003e-05,
+      "loss": 1.7027,
+      "step": 13271
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.70018405452464e-05,
+      "loss": 2.1891,
+      "step": 13272
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.700141246689784e-05,
+      "loss": 2.0096,
+      "step": 13273
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.7000984363380856e-05,
+      "loss": 1.7826,
+      "step": 13274
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.7000556234696984e-05,
+      "loss": 1.7485,
+      "step": 13275
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7000128080847777e-05,
+      "loss": 1.914,
+      "step": 13276
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6999699901834756e-05,
+      "loss": 1.79,
+      "step": 13277
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6999271697659475e-05,
+      "loss": 2.1805,
+      "step": 13278
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6998843468323464e-05,
+      "loss": 1.8091,
+      "step": 13279
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.699841521382827e-05,
+      "loss": 2.0715,
+      "step": 13280
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6997986934175423e-05,
+      "loss": 2.0037,
+      "step": 13281
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.699755862936647e-05,
+      "loss": 1.7747,
+      "step": 13282
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.699713029940295e-05,
+      "loss": 1.9156,
+      "step": 13283
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6996701944286403e-05,
+      "loss": 2.3027,
+      "step": 13284
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6996273564018363e-05,
+      "loss": 1.7776,
+      "step": 13285
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6995845158600377e-05,
+      "loss": 1.882,
+      "step": 13286
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.699541672803398e-05,
+      "loss": 1.7337,
+      "step": 13287
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6994988272320716e-05,
+      "loss": 1.871,
+      "step": 13288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.699455979146212e-05,
+      "loss": 1.6659,
+      "step": 13289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6994131285459744e-05,
+      "loss": 1.8947,
+      "step": 13290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.699370275431511e-05,
+      "loss": 1.824,
+      "step": 13291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6993274198029775e-05,
+      "loss": 1.8955,
+      "step": 13292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.699284561660527e-05,
+      "loss": 1.9218,
+      "step": 13293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6992417010043144e-05,
+      "loss": 1.934,
+      "step": 13294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6991988378344925e-05,
+      "loss": 1.7733,
+      "step": 13295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.6991559721512168e-05,
+      "loss": 2.6289,
+      "step": 13296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6991131039546402e-05,
+      "loss": 1.956,
+      "step": 13297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6990702332449175e-05,
+      "loss": 2.0197,
+      "step": 13298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.699027360022203e-05,
+      "loss": 2.987,
+      "step": 13299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6989844842866502e-05,
+      "loss": 2.0515,
+      "step": 13300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6989416060384133e-05,
+      "loss": 1.9172,
+      "step": 13301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.698898725277647e-05,
+      "loss": 2.2441,
+      "step": 13302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6988558420045045e-05,
+      "loss": 2.039,
+      "step": 13303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6988129562191407e-05,
+      "loss": 1.8431,
+      "step": 13304
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.69877006792171e-05,
+      "loss": 1.8778,
+      "step": 13305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6987271771123657e-05,
+      "loss": 2.1841,
+      "step": 13306
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6986842837912628e-05,
+      "loss": 2.2633,
+      "step": 13307
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.6986413879585548e-05,
+      "loss": 2.3703,
+      "step": 13308
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6985984896143964e-05,
+      "loss": 1.9695,
+      "step": 13309
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6985555887589416e-05,
+      "loss": 2.0009,
+      "step": 13310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6985126853923447e-05,
+      "loss": 2.2456,
+      "step": 13311
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.69846977951476e-05,
+      "loss": 1.5978,
+      "step": 13312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0,
+      "learning_rate": 1.6984268711263414e-05,
+      "loss": 2.0133,
+      "step": 13313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6983839602272435e-05,
+      "loss": 1.9136,
+      "step": 13314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6983410468176206e-05,
+      "loss": 2.0444,
+      "step": 13315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6982981308976265e-05,
+      "loss": 1.9273,
+      "step": 13316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6982552124674157e-05,
+      "loss": 2.2471,
+      "step": 13317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6982122915271427e-05,
+      "loss": 1.8355,
+      "step": 13318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6981693680769617e-05,
+      "loss": 1.8258,
+      "step": 13319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6981264421170268e-05,
+      "loss": 1.8814,
+      "step": 13320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.698083513647493e-05,
+      "loss": 1.9207,
+      "step": 13321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6980405826685133e-05,
+      "loss": 1.9672,
+      "step": 13322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.697997649180243e-05,
+      "loss": 1.9555,
+      "step": 13323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6979547131828363e-05,
+      "loss": 3.111,
+      "step": 13324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6979117746764475e-05,
+      "loss": 2.2669,
+      "step": 13325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.697868833661231e-05,
+      "loss": 2.121,
+      "step": 13326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.697825890137341e-05,
+      "loss": 1.9116,
+      "step": 13327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.697782944104932e-05,
+      "loss": 2.1333,
+      "step": 13328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.697739995564158e-05,
+      "loss": 1.9166,
+      "step": 13329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6976970445151743e-05,
+      "loss": 1.6284,
+      "step": 13330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.697654090958134e-05,
+      "loss": 1.9914,
+      "step": 13331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6976111348931928e-05,
+      "loss": 1.7643,
+      "step": 13332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6975681763205046e-05,
+      "loss": 1.7957,
+      "step": 13333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6975252152402232e-05,
+      "loss": 2.0049,
+      "step": 13334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6974822516525044e-05,
+      "loss": 1.7338,
+      "step": 13335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.697439285557501e-05,
+      "loss": 1.6315,
+      "step": 13336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.697396316955369e-05,
+      "loss": 1.476,
+      "step": 13337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6973533458462617e-05,
+      "loss": 1.943,
+      "step": 13338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6973103722303344e-05,
+      "loss": 1.6375,
+      "step": 13339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.697267396107741e-05,
+      "loss": 1.7977,
+      "step": 13340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6972244174786367e-05,
+      "loss": 1.6566,
+      "step": 13341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.697181436343175e-05,
+      "loss": 1.961,
+      "step": 13342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.697138452701511e-05,
+      "loss": 1.9123,
+      "step": 13343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.697095466553799e-05,
+      "loss": 2.0889,
+      "step": 13344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6970524779001936e-05,
+      "loss": 1.5913,
+      "step": 13345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6970094867408497e-05,
+      "loss": 1.6059,
+      "step": 13346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.6969664930759213e-05,
+      "loss": 2.5703,
+      "step": 13347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6969234969055636e-05,
+      "loss": 1.8792,
+      "step": 13348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.69688049822993e-05,
+      "loss": 1.8646,
+      "step": 13349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6968374970491764e-05,
+      "loss": 1.8889,
+      "step": 13350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.696794493363457e-05,
+      "loss": 1.7426,
+      "step": 13351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6967514871729255e-05,
+      "loss": 1.5406,
+      "step": 13352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6967084784777376e-05,
+      "loss": 1.8462,
+      "step": 13353
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6966654672780474e-05,
+      "loss": 1.6981,
+      "step": 13354
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6966224535740095e-05,
+      "loss": 1.8388,
+      "step": 13355
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.696579437365779e-05,
+      "loss": 2.2747,
+      "step": 13356
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6965364186535097e-05,
+      "loss": 1.8934,
+      "step": 13357
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.696493397437357e-05,
+      "loss": 1.6666,
+      "step": 13358
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6964503737174754e-05,
+      "loss": 1.8812,
+      "step": 13359
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.696407347494019e-05,
+      "loss": 1.6253,
+      "step": 13360
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.696364318767143e-05,
+      "loss": 2.0864,
+      "step": 13361
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.696321287537002e-05,
+      "loss": 1.9215,
+      "step": 13362
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6962782538037507e-05,
+      "loss": 1.7779,
+      "step": 13363
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6962352175675437e-05,
+      "loss": 1.637,
+      "step": 13364
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.696192178828536e-05,
+      "loss": 1.8758,
+      "step": 13365
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6961491375868816e-05,
+      "loss": 2.0891,
+      "step": 13366
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6961060938427358e-05,
+      "loss": 1.9225,
+      "step": 13367
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.6960630475962535e-05,
+      "loss": 1.8308,
+      "step": 13368
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.696019998847589e-05,
+      "loss": 1.8384,
+      "step": 13369
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.695976947596897e-05,
+      "loss": 2.1194,
+      "step": 13370
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6959338938443326e-05,
+      "loss": 2.0558,
+      "step": 13371
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6958908375900506e-05,
+      "loss": 1.9981,
+      "step": 13372
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6958477788342052e-05,
+      "loss": 1.7892,
+      "step": 13373
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6958047175769522e-05,
+      "loss": 1.7431,
+      "step": 13374
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6957616538184452e-05,
+      "loss": 2.0432,
+      "step": 13375
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.6957185875588403e-05,
+      "loss": 1.6797,
+      "step": 13376
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6956755187982913e-05,
+      "loss": 2.0067,
+      "step": 13377
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.695632447536953e-05,
+      "loss": 2.0872,
+      "step": 13378
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.6955893737749806e-05,
+      "loss": 1.6145,
+      "step": 13379
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.695546297512529e-05,
+      "loss": 1.7053,
+      "step": 13380
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6955032187497533e-05,
+      "loss": 1.8705,
+      "step": 13381
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6954601374868076e-05,
+      "loss": 1.7753,
+      "step": 13382
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6954170537238476e-05,
+      "loss": 1.9142,
+      "step": 13383
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6953739674610273e-05,
+      "loss": 1.7431,
+      "step": 13384
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6953308786985024e-05,
+      "loss": 1.894,
+      "step": 13385
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.695287787436427e-05,
+      "loss": 1.721,
+      "step": 13386
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.695244693674957e-05,
+      "loss": 1.9697,
+      "step": 13387
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6952015974142463e-05,
+      "loss": 1.9074,
+      "step": 13388
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6951584986544503e-05,
+      "loss": 2.0771,
+      "step": 13389
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.695115397395724e-05,
+      "loss": 1.9863,
+      "step": 13390
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6950722936382225e-05,
+      "loss": 2.0152,
+      "step": 13391
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6950291873821003e-05,
+      "loss": 1.6598,
+      "step": 13392
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6949860786275124e-05,
+      "loss": 1.7347,
+      "step": 13393
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6949429673746137e-05,
+      "loss": 1.9655,
+      "step": 13394
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6948998536235598e-05,
+      "loss": 1.9654,
+      "step": 13395
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.694856737374505e-05,
+      "loss": 1.8867,
+      "step": 13396
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6948136186276047e-05,
+      "loss": 1.6734,
+      "step": 13397
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.694770497383014e-05,
+      "loss": 1.9076,
+      "step": 13398
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.694727373640887e-05,
+      "loss": 1.8185,
+      "step": 13399
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.69468424740138e-05,
+      "loss": 1.8483,
+      "step": 13400
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6946411186646472e-05,
+      "loss": 1.7307,
+      "step": 13401
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6945979874308437e-05,
+      "loss": 1.9594,
+      "step": 13402
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.694554853700125e-05,
+      "loss": 1.9053,
+      "step": 13403
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.694511717472646e-05,
+      "loss": 1.6293,
+      "step": 13404
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6944685787485612e-05,
+      "loss": 2.1041,
+      "step": 13405
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6944254375280262e-05,
+      "loss": 1.8103,
+      "step": 13406
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.694382293811196e-05,
+      "loss": 1.8979,
+      "step": 13407
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6943391475982257e-05,
+      "loss": 1.9628,
+      "step": 13408
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6942959988892705e-05,
+      "loss": 1.7922,
+      "step": 13409
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.694252847684485e-05,
+      "loss": 1.8133,
+      "step": 13410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.694209693984025e-05,
+      "loss": 2.2343,
+      "step": 13411
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6941665377880454e-05,
+      "loss": 2.0995,
+      "step": 13412
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.694123379096701e-05,
+      "loss": 2.0044,
+      "step": 13413
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6940802179101475e-05,
+      "loss": 1.6477,
+      "step": 13414
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6940370542285396e-05,
+      "loss": 2.0299,
+      "step": 13415
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6939938880520325e-05,
+      "loss": 1.7651,
+      "step": 13416
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6939507193807818e-05,
+      "loss": 1.8179,
+      "step": 13417
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6939075482149422e-05,
+      "loss": 1.7952,
+      "step": 13418
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.693864374554669e-05,
+      "loss": 2.2378,
+      "step": 13419
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6938211984001177e-05,
+      "loss": 1.8527,
+      "step": 13420
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.693778019751443e-05,
+      "loss": 2.1528,
+      "step": 13421
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6937348386088005e-05,
+      "loss": 1.848,
+      "step": 13422
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6936916549723452e-05,
+      "loss": 1.7718,
+      "step": 13423
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6936484688422327e-05,
+      "loss": 2.0115,
+      "step": 13424
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.693605280218618e-05,
+      "loss": 1.8393,
+      "step": 13425
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6935620891016562e-05,
+      "loss": 1.6951,
+      "step": 13426
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6935188954915028e-05,
+      "loss": 2.1755,
+      "step": 13427
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.693475699388313e-05,
+      "loss": 1.6994,
+      "step": 13428
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6934325007922418e-05,
+      "loss": 2.1789,
+      "step": 13429
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.693389299703445e-05,
+      "loss": 1.9047,
+      "step": 13430
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6933460961220775e-05,
+      "loss": 2.0035,
+      "step": 13431
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.693302890048295e-05,
+      "loss": 1.9964,
+      "step": 13432
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6932596814822524e-05,
+      "loss": 1.943,
+      "step": 13433
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6932164704241053e-05,
+      "loss": 2.0984,
+      "step": 13434
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6931732568740086e-05,
+      "loss": 1.6919,
+      "step": 13435
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6931300408321182e-05,
+      "loss": 1.8277,
+      "step": 13436
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6930868222985896e-05,
+      "loss": 1.5855,
+      "step": 13437
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6930436012735772e-05,
+      "loss": 1.9311,
+      "step": 13438
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6930003777572373e-05,
+      "loss": 1.9186,
+      "step": 13439
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6929571517497248e-05,
+      "loss": 1.759,
+      "step": 13440
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6929139232511954e-05,
+      "loss": 2.0739,
+      "step": 13441
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.692870692261804e-05,
+      "loss": 1.7445,
+      "step": 13442
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6928274587817065e-05,
+      "loss": 1.9203,
+      "step": 13443
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.6927842228110583e-05,
+      "loss": 1.8601,
+      "step": 13444
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6927409843500143e-05,
+      "loss": 2.1208,
+      "step": 13445
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6926977433987305e-05,
+      "loss": 1.6981,
+      "step": 13446
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6926544999573618e-05,
+      "loss": 1.6459,
+      "step": 13447
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6926112540260646e-05,
+      "loss": 1.9721,
+      "step": 13448
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.692568005604993e-05,
+      "loss": 1.8827,
+      "step": 13449
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.692524754694304e-05,
+      "loss": 2.3491,
+      "step": 13450
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6924815012941517e-05,
+      "loss": 1.9048,
+      "step": 13451
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6924382454046923e-05,
+      "loss": 1.8077,
+      "step": 13452
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6923949870260813e-05,
+      "loss": 1.8429,
+      "step": 13453
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6923517261584745e-05,
+      "loss": 2.4837,
+      "step": 13454
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.692308462802026e-05,
+      "loss": 1.8118,
+      "step": 13455
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6922651969568927e-05,
+      "loss": 2.1582,
+      "step": 13456
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6922219286232298e-05,
+      "loss": 1.764,
+      "step": 13457
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.692178657801193e-05,
+      "loss": 1.8014,
+      "step": 13458
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6921353844909373e-05,
+      "loss": 2.1057,
+      "step": 13459
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6920921086926187e-05,
+      "loss": 1.9741,
+      "step": 13460
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.692048830406393e-05,
+      "loss": 1.6362,
+      "step": 13461
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6920055496324147e-05,
+      "loss": 1.6869,
+      "step": 13462
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6919622663708404e-05,
+      "loss": 2.0661,
+      "step": 13463
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.6919189806218257e-05,
+      "loss": 2.0071,
+      "step": 13464
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6918756923855255e-05,
+      "loss": 2.0557,
+      "step": 13465
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.691832401662096e-05,
+      "loss": 1.5023,
+      "step": 13466
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6917891084516925e-05,
+      "loss": 2.0857,
+      "step": 13467
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.691745812754471e-05,
+      "loss": 1.8838,
+      "step": 13468
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6917025145705868e-05,
+      "loss": 1.9221,
+      "step": 13469
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6916592139001954e-05,
+      "loss": 1.9545,
+      "step": 13470
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.691615910743453e-05,
+      "loss": 1.8022,
+      "step": 13471
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.691572605100515e-05,
+      "loss": 1.9155,
+      "step": 13472
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.691529296971537e-05,
+      "loss": 2.2495,
+      "step": 13473
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6914859863566743e-05,
+      "loss": 1.7915,
+      "step": 13474
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6914426732560835e-05,
+      "loss": 2.108,
+      "step": 13475
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6913993576699196e-05,
+      "loss": 2.0938,
+      "step": 13476
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6913560395983384e-05,
+      "loss": 1.6596,
+      "step": 13477
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.691312719041496e-05,
+      "loss": 1.8134,
+      "step": 13478
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6912693959995476e-05,
+      "loss": 2.0106,
+      "step": 13479
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6912260704726494e-05,
+      "loss": 1.7704,
+      "step": 13480
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.691182742460957e-05,
+      "loss": 1.6085,
+      "step": 13481
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.691139411964626e-05,
+      "loss": 1.7885,
+      "step": 13482
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6910960789838123e-05,
+      "loss": 1.8921,
+      "step": 13483
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6910527435186715e-05,
+      "loss": 1.931,
+      "step": 13484
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.6910094055693598e-05,
+      "loss": 1.8565,
+      "step": 13485
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6909660651360322e-05,
+      "loss": 1.9073,
+      "step": 13486
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6909227222188456e-05,
+      "loss": 1.7577,
+      "step": 13487
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.690879376817955e-05,
+      "loss": 1.8894,
+      "step": 13488
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.690836028933516e-05,
+      "loss": 2.1524,
+      "step": 13489
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6907926785656854e-05,
+      "loss": 1.7939,
+      "step": 13490
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6907493257146183e-05,
+      "loss": 1.8926,
+      "step": 13491
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6907059703804708e-05,
+      "loss": 1.6233,
+      "step": 13492
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6906626125633986e-05,
+      "loss": 1.8398,
+      "step": 13493
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6906192522635577e-05,
+      "loss": 1.9275,
+      "step": 13494
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.690575889481104e-05,
+      "loss": 1.6564,
+      "step": 13495
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6905325242161933e-05,
+      "loss": 2.0557,
+      "step": 13496
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6904891564689812e-05,
+      "loss": 1.904,
+      "step": 13497
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.690445786239624e-05,
+      "loss": 1.8107,
+      "step": 13498
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.6904024135282778e-05,
+      "loss": 1.8902,
+      "step": 13499
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6903590383350975e-05,
+      "loss": 2.051,
+      "step": 13500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6903156606602405e-05,
+      "loss": 1.86,
+      "step": 13501
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6902722805038612e-05,
+      "loss": 1.6683,
+      "step": 13502
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.690228897866117e-05,
+      "loss": 1.9035,
+      "step": 13503
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6901855127471627e-05,
+      "loss": 1.9778,
+      "step": 13504
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.690142125147155e-05,
+      "loss": 1.718,
+      "step": 13505
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.6900987350662493e-05,
+      "loss": 1.8782,
+      "step": 13506
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.690055342504602e-05,
+      "loss": 1.9414,
+      "step": 13507
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.690011947462369e-05,
+      "loss": 1.7593,
+      "step": 13508
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6899685499397062e-05,
+      "loss": 1.7644,
+      "step": 13509
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6899251499367693e-05,
+      "loss": 1.4862,
+      "step": 13510
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6898817474537152e-05,
+      "loss": 1.8328,
+      "step": 13511
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.689838342490699e-05,
+      "loss": 1.8267,
+      "step": 13512
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6897949350478774e-05,
+      "loss": 1.8518,
+      "step": 13513
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6897515251254058e-05,
+      "loss": 2.0611,
+      "step": 13514
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.689708112723441e-05,
+      "loss": 2.0568,
+      "step": 13515
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.689664697842138e-05,
+      "loss": 2.0785,
+      "step": 13516
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.689621280481654e-05,
+      "loss": 1.8661,
+      "step": 13517
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6895778606421446e-05,
+      "loss": 2.0633,
+      "step": 13518
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.6895344383237656e-05,
+      "loss": 1.9601,
+      "step": 13519
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6894910135266734e-05,
+      "loss": 1.8923,
+      "step": 13520
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.6894475862510242e-05,
+      "loss": 2.1868,
+      "step": 13521
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6894041564969743e-05,
+      "loss": 2.2516,
+      "step": 13522
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.689360724264679e-05,
+      "loss": 1.91,
+      "step": 13523
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.689317289554295e-05,
+      "loss": 1.987,
+      "step": 13524
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6892738523659785e-05,
+      "loss": 2.032,
+      "step": 13525
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6892304126998857e-05,
+      "loss": 2.1077,
+      "step": 13526
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.689186970556172e-05,
+      "loss": 1.6999,
+      "step": 13527
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6891435259349946e-05,
+      "loss": 2.4154,
+      "step": 13528
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6891000788365092e-05,
+      "loss": 1.7611,
+      "step": 13529
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6890566292608715e-05,
+      "loss": 1.8742,
+      "step": 13530
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6890131772082384e-05,
+      "loss": 2.1256,
+      "step": 13531
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6889697226787665e-05,
+      "loss": 1.8859,
+      "step": 13532
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.6889262656726105e-05,
+      "loss": 1.7809,
+      "step": 13533
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6888828061899277e-05,
+      "loss": 1.7805,
+      "step": 13534
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6888393442308747e-05,
+      "loss": 1.6801,
+      "step": 13535
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6887958797956065e-05,
+      "loss": 1.5522,
+      "step": 13536
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6887524128842805e-05,
+      "loss": 1.8306,
+      "step": 13537
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6887089434970524e-05,
+      "loss": 1.6098,
+      "step": 13538
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.688665471634078e-05,
+      "loss": 1.8469,
+      "step": 13539
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.688621997295515e-05,
+      "loss": 1.7318,
+      "step": 13540
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6885785204815178e-05,
+      "loss": 1.8567,
+      "step": 13541
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.688535041192244e-05,
+      "loss": 1.8129,
+      "step": 13542
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6884915594278498e-05,
+      "loss": 1.9429,
+      "step": 13543
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6884480751884912e-05,
+      "loss": 1.8581,
+      "step": 13544
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6884045884743243e-05,
+      "loss": 1.8731,
+      "step": 13545
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.688361099285506e-05,
+      "loss": 1.8084,
+      "step": 13546
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6883176076221922e-05,
+      "loss": 1.9751,
+      "step": 13547
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6882741134845394e-05,
+      "loss": 2.071,
+      "step": 13548
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.688230616872704e-05,
+      "loss": 2.0636,
+      "step": 13549
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6881871177868424e-05,
+      "loss": 1.9983,
+      "step": 13550
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6881436162271105e-05,
+      "loss": 2.2271,
+      "step": 13551
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6881001121936657e-05,
+      "loss": 2.1728,
+      "step": 13552
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.688056605686663e-05,
+      "loss": 1.9793,
+      "step": 13553
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6880130967062598e-05,
+      "loss": 1.892,
+      "step": 13554
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6879695852526122e-05,
+      "loss": 1.8229,
+      "step": 13555
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6879260713258764e-05,
+      "loss": 1.8743,
+      "step": 13556
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.6878825549262096e-05,
+      "loss": 2.0068,
+      "step": 13557
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.6878390360537677e-05,
+      "loss": 1.5789,
+      "step": 13558
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6877955147087065e-05,
+      "loss": 1.7733,
+      "step": 13559
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6877519908911836e-05,
+      "loss": 1.8225,
+      "step": 13560
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6877084646013548e-05,
+      "loss": 2.0105,
+      "step": 13561
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6876649358393766e-05,
+      "loss": 2.0336,
+      "step": 13562
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6876214046054057e-05,
+      "loss": 2.1964,
+      "step": 13563
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6875778708995984e-05,
+      "loss": 1.7275,
+      "step": 13564
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.6875343347221114e-05,
+      "loss": 1.6907,
+      "step": 13565
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.687490796073101e-05,
+      "loss": 1.9031,
+      "step": 13566
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.6874472549527237e-05,
+      "loss": 1.5626,
+      "step": 13567
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.687403711361136e-05,
+      "loss": 1.714,
+      "step": 13568
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.687360165298495e-05,
+      "loss": 2.1875,
+      "step": 13569
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6873166167649562e-05,
+      "loss": 1.6331,
+      "step": 13570
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6872730657606772e-05,
+      "loss": 2.0365,
+      "step": 13571
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.687229512285814e-05,
+      "loss": 2.2754,
+      "step": 13572
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.687185956340523e-05,
+      "loss": 2.097,
+      "step": 13573
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.687142397924961e-05,
+      "loss": 1.9068,
+      "step": 13574
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6870988370392846e-05,
+      "loss": 1.9209,
+      "step": 13575
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.68705527368365e-05,
+      "loss": 1.7823,
+      "step": 13576
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.687011707858215e-05,
+      "loss": 2.0647,
+      "step": 13577
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6869681395631346e-05,
+      "loss": 2.1314,
+      "step": 13578
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.686924568798567e-05,
+      "loss": 1.8157,
+      "step": 13579
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6868809955646672e-05,
+      "loss": 1.9028,
+      "step": 13580
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6868374198615933e-05,
+      "loss": 2.1648,
+      "step": 13581
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6867938416895013e-05,
+      "loss": 1.8983,
+      "step": 13582
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6867502610485474e-05,
+      "loss": 2.0164,
+      "step": 13583
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6867066779388886e-05,
+      "loss": 1.6091,
+      "step": 13584
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.375,
+      "learning_rate": 1.686663092360682e-05,
+      "loss": 1.9482,
+      "step": 13585
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.6866195043140842e-05,
+      "loss": 1.9616,
+      "step": 13586
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6865759137992514e-05,
+      "loss": 1.9922,
+      "step": 13587
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6865323208163408e-05,
+      "loss": 1.9299,
+      "step": 13588
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6864887253655088e-05,
+      "loss": 2.0306,
+      "step": 13589
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.686445127446912e-05,
+      "loss": 1.7959,
+      "step": 13590
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.686401527060707e-05,
+      "loss": 2.437,
+      "step": 13591
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6863579242070515e-05,
+      "loss": 1.8319,
+      "step": 13592
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6863143188861014e-05,
+      "loss": 2.1403,
+      "step": 13593
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6862707110980133e-05,
+      "loss": 1.8137,
+      "step": 13594
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.686227100842945e-05,
+      "loss": 1.9884,
+      "step": 13595
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6861834881210518e-05,
+      "loss": 2.088,
+      "step": 13596
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6861398729324918e-05,
+      "loss": 1.7721,
+      "step": 13597
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6860962552774206e-05,
+      "loss": 2.0059,
+      "step": 13598
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6860526351559964e-05,
+      "loss": 1.9902,
+      "step": 13599
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.686009012568375e-05,
+      "loss": 1.9712,
+      "step": 13600
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.6859653875147128e-05,
+      "loss": 1.9675,
+      "step": 13601
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6859217599951677e-05,
+      "loss": 2.1661,
+      "step": 13602
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6858781300098962e-05,
+      "loss": 1.8482,
+      "step": 13603
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6858344975590547e-05,
+      "loss": 1.6667,
+      "step": 13604
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.685790862642801e-05,
+      "loss": 1.4367,
+      "step": 13605
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.685747225261291e-05,
+      "loss": 1.8996,
+      "step": 13606
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6857035854146815e-05,
+      "loss": 1.9078,
+      "step": 13607
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.68565994310313e-05,
+      "loss": 1.7332,
+      "step": 13608
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6856162983267937e-05,
+      "loss": 1.6795,
+      "step": 13609
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6855726510858282e-05,
+      "loss": 1.584,
+      "step": 13610
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.6855290013803913e-05,
+      "loss": 1.9847,
+      "step": 13611
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.68548534921064e-05,
+      "loss": 1.8475,
+      "step": 13612
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.685441694576731e-05,
+      "loss": 1.7966,
+      "step": 13613
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6853980374788213e-05,
+      "loss": 1.7608,
+      "step": 13614
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.6853543779170675e-05,
+      "loss": 2.0862,
+      "step": 13615
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6853107158916268e-05,
+      "loss": 1.8518,
+      "step": 13616
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6852670514026562e-05,
+      "loss": 2.0968,
+      "step": 13617
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6852233844503125e-05,
+      "loss": 1.7934,
+      "step": 13618
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.685179715034753e-05,
+      "loss": 1.7433,
+      "step": 13619
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.6851360431561342e-05,
+      "loss": 2.1306,
+      "step": 13620
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6850923688146135e-05,
+      "loss": 1.93,
+      "step": 13621
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.685048692010348e-05,
+      "loss": 1.9417,
+      "step": 13622
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.685005012743494e-05,
+      "loss": 2.4935,
+      "step": 13623
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6849613310142094e-05,
+      "loss": 1.7426,
+      "step": 13624
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6849176468226507e-05,
+      "loss": 1.919,
+      "step": 13625
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.684873960168975e-05,
+      "loss": 1.9209,
+      "step": 13626
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6848302710533393e-05,
+      "loss": 1.9361,
+      "step": 13627
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6847865794759006e-05,
+      "loss": 1.8785,
+      "step": 13628
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6847428854368165e-05,
+      "loss": 1.8574,
+      "step": 13629
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.6846991889362435e-05,
+      "loss": 1.9618,
+      "step": 13630
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6846554899743387e-05,
+      "loss": 1.846,
+      "step": 13631
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6846117885512595e-05,
+      "loss": 1.7561,
+      "step": 13632
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6845680846671626e-05,
+      "loss": 1.9482,
+      "step": 13633
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.6845243783222052e-05,
+      "loss": 1.8581,
+      "step": 13634
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.684480669516545e-05,
+      "loss": 1.6385,
+      "step": 13635
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6844369582503385e-05,
+      "loss": 1.8261,
+      "step": 13636
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.684393244523743e-05,
+      "loss": 1.7855,
+      "step": 13637
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6843495283369155e-05,
+      "loss": 2.1354,
+      "step": 13638
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6843058096900135e-05,
+      "loss": 1.868,
+      "step": 13639
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6842620885831937e-05,
+      "loss": 1.8614,
+      "step": 13640
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.684218365016614e-05,
+      "loss": 1.9067,
+      "step": 13641
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6841746389904306e-05,
+      "loss": 1.9039,
+      "step": 13642
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6841309105048015e-05,
+      "loss": 2.2267,
+      "step": 13643
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.375,
+      "learning_rate": 1.684087179559883e-05,
+      "loss": 1.9484,
+      "step": 13644
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6840434461558335e-05,
+      "loss": 1.9547,
+      "step": 13645
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6839997102928093e-05,
+      "loss": 1.8194,
+      "step": 13646
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.683955971970968e-05,
+      "loss": 1.8476,
+      "step": 13647
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6839122311904666e-05,
+      "loss": 1.8734,
+      "step": 13648
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.683868487951463e-05,
+      "loss": 1.8937,
+      "step": 13649
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.683824742254113e-05,
+      "loss": 1.7902,
+      "step": 13650
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6837809940985755e-05,
+      "loss": 1.8778,
+      "step": 13651
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.683737243485007e-05,
+      "loss": 1.8517,
+      "step": 13652
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6836934904135644e-05,
+      "loss": 2.0387,
+      "step": 13653
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6836497348844054e-05,
+      "loss": 2.0415,
+      "step": 13654
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6836059768976877e-05,
+      "loss": 1.6713,
+      "step": 13655
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6835622164535678e-05,
+      "loss": 1.8248,
+      "step": 13656
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.6835184535522036e-05,
+      "loss": 1.8196,
+      "step": 13657
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.683474688193752e-05,
+      "loss": 2.1583,
+      "step": 13658
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.683430920378371e-05,
+      "loss": 2.0165,
+      "step": 13659
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.683387150106217e-05,
+      "loss": 1.8614,
+      "step": 13660
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6833433773774478e-05,
+      "loss": 2.0089,
+      "step": 13661
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.683299602192221e-05,
+      "loss": 2.0349,
+      "step": 13662
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6832558245506937e-05,
+      "loss": 1.9042,
+      "step": 13663
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.6832120444530232e-05,
+      "loss": 2.0264,
+      "step": 13664
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.683168261899367e-05,
+      "loss": 1.5897,
+      "step": 13665
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6831244768898826e-05,
+      "loss": 1.8158,
+      "step": 13666
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.683080689424727e-05,
+      "loss": 2.9952,
+      "step": 13667
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.683036899504058e-05,
+      "loss": 2.0826,
+      "step": 13668
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.682993107128033e-05,
+      "loss": 1.665,
+      "step": 13669
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6829493122968092e-05,
+      "loss": 2.1981,
+      "step": 13670
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.682905515010544e-05,
+      "loss": 1.9725,
+      "step": 13671
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6828617152693952e-05,
+      "loss": 1.9005,
+      "step": 13672
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.68281791307352e-05,
+      "loss": 2.0101,
+      "step": 13673
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.682774108423076e-05,
+      "loss": 1.5685,
+      "step": 13674
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.68273030131822e-05,
+      "loss": 1.9543,
+      "step": 13675
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6826864917591105e-05,
+      "loss": 2.1626,
+      "step": 13676
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6826426797459046e-05,
+      "loss": 1.6892,
+      "step": 13677
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6825988652787593e-05,
+      "loss": 1.7881,
+      "step": 13678
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6825550483578327e-05,
+      "loss": 1.777,
+      "step": 13679
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.682511228983282e-05,
+      "loss": 1.9797,
+      "step": 13680
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.682467407155265e-05,
+      "loss": 1.844,
+      "step": 13681
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.682423582873939e-05,
+      "loss": 1.745,
+      "step": 13682
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6823797561394618e-05,
+      "loss": 1.8899,
+      "step": 13683
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6823359269519904e-05,
+      "loss": 2.2681,
+      "step": 13684
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.682292095311683e-05,
+      "loss": 2.0874,
+      "step": 13685
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.682248261218697e-05,
+      "loss": 1.6825,
+      "step": 13686
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6822044246731898e-05,
+      "loss": 1.7173,
+      "step": 13687
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6821605856753188e-05,
+      "loss": 1.7774,
+      "step": 13688
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6821167442252416e-05,
+      "loss": 2.1045,
+      "step": 13689
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6820729003231165e-05,
+      "loss": 2.015,
+      "step": 13690
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6820290539691005e-05,
+      "loss": 1.8686,
+      "step": 13691
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6819852051633512e-05,
+      "loss": 1.7702,
+      "step": 13692
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6819413539060264e-05,
+      "loss": 1.7017,
+      "step": 13693
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.681897500197284e-05,
+      "loss": 1.823,
+      "step": 13694
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.681853644037281e-05,
+      "loss": 2.3071,
+      "step": 13695
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6818097854261756e-05,
+      "loss": 1.913,
+      "step": 13696
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6817659243641253e-05,
+      "loss": 2.1873,
+      "step": 13697
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6817220608512874e-05,
+      "loss": 1.6211,
+      "step": 13698
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6816781948878203e-05,
+      "loss": 1.9746,
+      "step": 13699
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6816343264738814e-05,
+      "loss": 1.6697,
+      "step": 13700
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.681590455609628e-05,
+      "loss": 1.6898,
+      "step": 13701
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.681546582295218e-05,
+      "loss": 1.5706,
+      "step": 13702
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.681502706530809e-05,
+      "loss": 1.8208,
+      "step": 13703
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6814588283165595e-05,
+      "loss": 1.8883,
+      "step": 13704
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6814149476526265e-05,
+      "loss": 1.6906,
+      "step": 13705
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.681371064539168e-05,
+      "loss": 1.7381,
+      "step": 13706
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6813271789763414e-05,
+      "loss": 2.2363,
+      "step": 13707
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6812832909643048e-05,
+      "loss": 1.7134,
+      "step": 13708
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.681239400503216e-05,
+      "loss": 1.5569,
+      "step": 13709
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6811955075932327e-05,
+      "loss": 1.6563,
+      "step": 13710
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6811516122345128e-05,
+      "loss": 1.9927,
+      "step": 13711
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6811077144272133e-05,
+      "loss": 1.7413,
+      "step": 13712
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6810638141714933e-05,
+      "loss": 1.8443,
+      "step": 13713
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.6810199114675096e-05,
+      "loss": 2.2677,
+      "step": 13714
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6809760063154204e-05,
+      "loss": 1.9212,
+      "step": 13715
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.680932098715384e-05,
+      "loss": 2.1417,
+      "step": 13716
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6808881886675572e-05,
+      "loss": 1.7623,
+      "step": 13717
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6808442761720986e-05,
+      "loss": 1.5848,
+      "step": 13718
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6808003612291658e-05,
+      "loss": 1.7748,
+      "step": 13719
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6807564438389166e-05,
+      "loss": 1.9948,
+      "step": 13720
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6807125240015092e-05,
+      "loss": 1.6792,
+      "step": 13721
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.680668601717101e-05,
+      "loss": 1.7886,
+      "step": 13722
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6806246769858502e-05,
+      "loss": 2.0712,
+      "step": 13723
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6805807498079148e-05,
+      "loss": 1.8241,
+      "step": 13724
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6805368201834523e-05,
+      "loss": 1.7806,
+      "step": 13725
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.680492888112621e-05,
+      "loss": 1.717,
+      "step": 13726
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.680448953595579e-05,
+      "loss": 1.9481,
+      "step": 13727
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6804050166324833e-05,
+      "loss": 1.8913,
+      "step": 13728
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.680361077223493e-05,
+      "loss": 1.7219,
+      "step": 13729
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6803171353687656e-05,
+      "loss": 1.5264,
+      "step": 13730
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6802731910684586e-05,
+      "loss": 1.8211,
+      "step": 13731
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6802292443227305e-05,
+      "loss": 1.8899,
+      "step": 13732
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6801852951317392e-05,
+      "loss": 1.9893,
+      "step": 13733
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6801413434956426e-05,
+      "loss": 2.3519,
+      "step": 13734
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6800973894145987e-05,
+      "loss": 1.8857,
+      "step": 13735
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6800534328887653e-05,
+      "loss": 1.8346,
+      "step": 13736
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6800094739183012e-05,
+      "loss": 1.9612,
+      "step": 13737
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6799655125033635e-05,
+      "loss": 2.0154,
+      "step": 13738
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6799215486441106e-05,
+      "loss": 2.0084,
+      "step": 13739
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6798775823407004e-05,
+      "loss": 1.8674,
+      "step": 13740
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6798336135932915e-05,
+      "loss": 1.7717,
+      "step": 13741
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.679789642402041e-05,
+      "loss": 1.8013,
+      "step": 13742
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.679745668767108e-05,
+      "loss": 1.838,
+      "step": 13743
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.67970169268865e-05,
+      "loss": 1.7666,
+      "step": 13744
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6796577141668247e-05,
+      "loss": 1.5936,
+      "step": 13745
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.679613733201791e-05,
+      "loss": 1.7471,
+      "step": 13746
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6795697497937067e-05,
+      "loss": 1.8806,
+      "step": 13747
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6795257639427297e-05,
+      "loss": 1.7747,
+      "step": 13748
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.679481775649018e-05,
+      "loss": 1.7722,
+      "step": 13749
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6794377849127304e-05,
+      "loss": 1.8689,
+      "step": 13750
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.679393791734025e-05,
+      "loss": 2.0374,
+      "step": 13751
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.679349796113059e-05,
+      "loss": 1.7231,
+      "step": 13752
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6793057980499914e-05,
+      "loss": 2.0282,
+      "step": 13753
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.67926179754498e-05,
+      "loss": 1.9543,
+      "step": 13754
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.679217794598183e-05,
+      "loss": 1.7276,
+      "step": 13755
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.679173789209759e-05,
+      "loss": 1.753,
+      "step": 13756
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6791297813798656e-05,
+      "loss": 1.588,
+      "step": 13757
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6790857711086613e-05,
+      "loss": 1.6113,
+      "step": 13758
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.679041758396304e-05,
+      "loss": 2.0306,
+      "step": 13759
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6789977432429524e-05,
+      "loss": 1.8135,
+      "step": 13760
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6789537256487648e-05,
+      "loss": 1.9668,
+      "step": 13761
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6789097056138986e-05,
+      "loss": 1.5196,
+      "step": 13762
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.678865683138513e-05,
+      "loss": 1.7063,
+      "step": 13763
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6788216582227656e-05,
+      "loss": 2.0955,
+      "step": 13764
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6787776308668147e-05,
+      "loss": 1.7959,
+      "step": 13765
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.678733601070819e-05,
+      "loss": 1.6316,
+      "step": 13766
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6786895688349364e-05,
+      "loss": 1.8072,
+      "step": 13767
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6786455341593252e-05,
+      "loss": 1.894,
+      "step": 13768
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6786014970441436e-05,
+      "loss": 2.0138,
+      "step": 13769
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6785574574895508e-05,
+      "loss": 1.7841,
+      "step": 13770
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.678513415495704e-05,
+      "loss": 1.5703,
+      "step": 13771
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6784693710627618e-05,
+      "loss": 1.9328,
+      "step": 13772
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.678425324190883e-05,
+      "loss": 2.0921,
+      "step": 13773
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6783812748802255e-05,
+      "loss": 1.8402,
+      "step": 13774
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6783372231309474e-05,
+      "loss": 1.7679,
+      "step": 13775
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.678293168943208e-05,
+      "loss": 1.9305,
+      "step": 13776
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6782491123171648e-05,
+      "loss": 1.9698,
+      "step": 13777
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6782050532529766e-05,
+      "loss": 1.9624,
+      "step": 13778
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6781609917508013e-05,
+      "loss": 1.7356,
+      "step": 13779
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.6781169278107977e-05,
+      "loss": 1.5983,
+      "step": 13780
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.678072861433124e-05,
+      "loss": 1.9151,
+      "step": 13781
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.678028792617939e-05,
+      "loss": 1.902,
+      "step": 13782
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.677984721365401e-05,
+      "loss": 1.9497,
+      "step": 13783
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.677940647675668e-05,
+      "loss": 1.5337,
+      "step": 13784
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6778965715488985e-05,
+      "loss": 1.6847,
+      "step": 13785
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6778524929852513e-05,
+      "loss": 1.96,
+      "step": 13786
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6778084119848847e-05,
+      "loss": 1.6656,
+      "step": 13787
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6777643285479574e-05,
+      "loss": 2.0649,
+      "step": 13788
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6777202426746276e-05,
+      "loss": 1.7851,
+      "step": 13789
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6776761543650535e-05,
+      "loss": 1.6904,
+      "step": 13790
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.677632063619394e-05,
+      "loss": 1.9247,
+      "step": 13791
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6775879704378076e-05,
+      "loss": 1.7688,
+      "step": 13792
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.6775438748204524e-05,
+      "loss": 1.7565,
+      "step": 13793
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6774997767674874e-05,
+      "loss": 1.9508,
+      "step": 13794
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.677455676279071e-05,
+      "loss": 1.8707,
+      "step": 13795
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6774115733553616e-05,
+      "loss": 1.7279,
+      "step": 13796
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.677367467996518e-05,
+      "loss": 1.569,
+      "step": 13797
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.677323360202698e-05,
+      "loss": 2.0449,
+      "step": 13798
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.677279249974061e-05,
+      "loss": 1.9898,
+      "step": 13799
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6772351373107657e-05,
+      "loss": 1.4989,
+      "step": 13800
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6771910222129696e-05,
+      "loss": 2.1663,
+      "step": 13801
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6771469046808324e-05,
+      "loss": 1.936,
+      "step": 13802
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6771027847145114e-05,
+      "loss": 1.6311,
+      "step": 13803
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6770586623141668e-05,
+      "loss": 1.8251,
+      "step": 13804
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6770145374799564e-05,
+      "loss": 1.9201,
+      "step": 13805
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6769704102120387e-05,
+      "loss": 2.1167,
+      "step": 13806
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6769262805105723e-05,
+      "loss": 2.0241,
+      "step": 13807
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6768821483757164e-05,
+      "loss": 1.8146,
+      "step": 13808
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.676838013807629e-05,
+      "loss": 2.0965,
+      "step": 13809
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.676793876806469e-05,
+      "loss": 1.885,
+      "step": 13810
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.676749737372395e-05,
+      "loss": 2.0331,
+      "step": 13811
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6767055955055656e-05,
+      "loss": 2.1391,
+      "step": 13812
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.67666145120614e-05,
+      "loss": 1.9044,
+      "step": 13813
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.6766173044742765e-05,
+      "loss": 1.8332,
+      "step": 13814
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6765731553101336e-05,
+      "loss": 1.7467,
+      "step": 13815
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.67652900371387e-05,
+      "loss": 2.1274,
+      "step": 13816
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6764848496856453e-05,
+      "loss": 1.9705,
+      "step": 13817
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.676440693225617e-05,
+      "loss": 1.8794,
+      "step": 13818
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.6763965343339447e-05,
+      "loss": 1.8727,
+      "step": 13819
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6763523730107865e-05,
+      "loss": 1.8625,
+      "step": 13820
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.6763082092563018e-05,
+      "loss": 2.086,
+      "step": 13821
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6762640430706486e-05,
+      "loss": 1.914,
+      "step": 13822
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.6762198744539865e-05,
+      "loss": 2.0718,
+      "step": 13823
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.676175703406474e-05,
+      "loss": 1.6548,
+      "step": 13824
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6761315299282693e-05,
+      "loss": 2.0802,
+      "step": 13825
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.676087354019532e-05,
+      "loss": 1.7506,
+      "step": 13826
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6760431756804204e-05,
+      "loss": 1.6761,
+      "step": 13827
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6759989949110933e-05,
+      "loss": 2.0933,
+      "step": 13828
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.67595481171171e-05,
+      "loss": 1.9984,
+      "step": 13829
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6759106260824287e-05,
+      "loss": 1.9576,
+      "step": 13830
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.675866438023409e-05,
+      "loss": 1.6072,
+      "step": 13831
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.675822247534809e-05,
+      "loss": 2.0773,
+      "step": 13832
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6757780546167874e-05,
+      "loss": 2.0193,
+      "step": 13833
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6757338592695042e-05,
+      "loss": 1.8033,
+      "step": 13834
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.6756896614931172e-05,
+      "loss": 1.9429,
+      "step": 13835
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6756454612877857e-05,
+      "loss": 1.9536,
+      "step": 13836
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.675601258653669e-05,
+      "loss": 1.894,
+      "step": 13837
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.6755570535909252e-05,
+      "loss": 1.7467,
+      "step": 13838
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6755128460997134e-05,
+      "loss": 2.0018,
+      "step": 13839
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6754686361801927e-05,
+      "loss": 1.5029,
+      "step": 13840
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6754244238325222e-05,
+      "loss": 2.1673,
+      "step": 13841
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6753802090568606e-05,
+      "loss": 2.0558,
+      "step": 13842
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6753359918533667e-05,
+      "loss": 1.7934,
+      "step": 13843
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6752917722222e-05,
+      "loss": 1.4916,
+      "step": 13844
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6752475501635186e-05,
+      "loss": 1.7914,
+      "step": 13845
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.675203325677482e-05,
+      "loss": 1.6888,
+      "step": 13846
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6751590987642497e-05,
+      "loss": 1.8524,
+      "step": 13847
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.6751148694239795e-05,
+      "loss": 3.8346,
+      "step": 13848
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6750706376568312e-05,
+      "loss": 1.7889,
+      "step": 13849
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6750264034629637e-05,
+      "loss": 1.6508,
+      "step": 13850
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.6749821668425358e-05,
+      "loss": 1.791,
+      "step": 13851
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.674937927795707e-05,
+      "loss": 1.882,
+      "step": 13852
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6748936863226354e-05,
+      "loss": 1.8959,
+      "step": 13853
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.674849442423481e-05,
+      "loss": 1.8617,
+      "step": 13854
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.674805196098402e-05,
+      "loss": 1.8554,
+      "step": 13855
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6747609473475585e-05,
+      "loss": 1.6578,
+      "step": 13856
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6747166961711086e-05,
+      "loss": 1.6982,
+      "step": 13857
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6746724425692117e-05,
+      "loss": 1.832,
+      "step": 13858
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.674628186542027e-05,
+      "loss": 1.7986,
+      "step": 13859
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6745839280897137e-05,
+      "loss": 1.7925,
+      "step": 13860
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6745396672124306e-05,
+      "loss": 2.0565,
+      "step": 13861
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6744954039103368e-05,
+      "loss": 1.6812,
+      "step": 13862
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6744511381835916e-05,
+      "loss": 1.7245,
+      "step": 13863
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6744068700323544e-05,
+      "loss": 1.8887,
+      "step": 13864
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6743625994567837e-05,
+      "loss": 2.2102,
+      "step": 13865
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6743183264570388e-05,
+      "loss": 1.9627,
+      "step": 13866
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.674274051033279e-05,
+      "loss": 2.1726,
+      "step": 13867
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.674229773185664e-05,
+      "loss": 2.1589,
+      "step": 13868
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.674185492914352e-05,
+      "loss": 1.9327,
+      "step": 13869
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.674141210219502e-05,
+      "loss": 1.8674,
+      "step": 13870
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6740969251012745e-05,
+      "loss": 2.0352,
+      "step": 13871
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.674052637559828e-05,
+      "loss": 1.9155,
+      "step": 13872
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.6740083475953216e-05,
+      "loss": 2.2072,
+      "step": 13873
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6739640552079145e-05,
+      "loss": 1.6398,
+      "step": 13874
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.673919760397766e-05,
+      "loss": 1.8796,
+      "step": 13875
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6738754631650352e-05,
+      "loss": 1.9083,
+      "step": 13876
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6738311635098818e-05,
+      "loss": 2.2669,
+      "step": 13877
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6737868614324645e-05,
+      "loss": 2.1104,
+      "step": 13878
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.673742556932943e-05,
+      "loss": 1.7761,
+      "step": 13879
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.673698250011476e-05,
+      "loss": 1.7694,
+      "step": 13880
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6736539406682234e-05,
+      "loss": 1.9989,
+      "step": 13881
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.673609628903344e-05,
+      "loss": 1.6532,
+      "step": 13882
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6735653147169973e-05,
+      "loss": 1.9068,
+      "step": 13883
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.6735209981093428e-05,
+      "loss": 1.7894,
+      "step": 13884
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.6734766790805394e-05,
+      "loss": 1.7728,
+      "step": 13885
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6734323576307464e-05,
+      "loss": 1.6453,
+      "step": 13886
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.673388033760124e-05,
+      "loss": 1.9583,
+      "step": 13887
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6733437074688305e-05,
+      "loss": 2.132,
+      "step": 13888
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6732993787570253e-05,
+      "loss": 1.9701,
+      "step": 13889
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6732550476248682e-05,
+      "loss": 1.7162,
+      "step": 13890
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6732107140725188e-05,
+      "loss": 1.856,
+      "step": 13891
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.673166378100136e-05,
+      "loss": 2.0146,
+      "step": 13892
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6731220397078788e-05,
+      "loss": 1.9673,
+      "step": 13893
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6730776988959078e-05,
+      "loss": 2.0898,
+      "step": 13894
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.673033355664381e-05,
+      "loss": 1.9337,
+      "step": 13895
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6729890100134588e-05,
+      "loss": 2.0675,
+      "step": 13896
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6729446619433e-05,
+      "loss": 1.9813,
+      "step": 13897
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.6729003114540644e-05,
+      "loss": 1.8801,
+      "step": 13898
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6728559585459113e-05,
+      "loss": 1.8245,
+      "step": 13899
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.6728116032190004e-05,
+      "loss": 1.566,
+      "step": 13900
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.672767245473491e-05,
+      "loss": 2.0248,
+      "step": 13901
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.672722885309542e-05,
+      "loss": 1.8783,
+      "step": 13902
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6726785227273135e-05,
+      "loss": 1.7204,
+      "step": 13903
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6726341577269645e-05,
+      "loss": 1.9368,
+      "step": 13904
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6725897903086556e-05,
+      "loss": 1.6195,
+      "step": 13905
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6725454204725443e-05,
+      "loss": 1.9613,
+      "step": 13906
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6725010482187922e-05,
+      "loss": 1.9905,
+      "step": 13907
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6724566735475577e-05,
+      "loss": 1.7422,
+      "step": 13908
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6724122964590002e-05,
+      "loss": 1.9246,
+      "step": 13909
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6723679169532797e-05,
+      "loss": 1.7258,
+      "step": 13910
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6723235350305554e-05,
+      "loss": 1.7865,
+      "step": 13911
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6722791506909873e-05,
+      "loss": 1.9418,
+      "step": 13912
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6722347639347342e-05,
+      "loss": 1.8481,
+      "step": 13913
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6721903747619568e-05,
+      "loss": 2.0072,
+      "step": 13914
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6721459831728136e-05,
+      "loss": 2.1129,
+      "step": 13915
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6721015891674642e-05,
+      "loss": 1.8694,
+      "step": 13916
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.672057192746069e-05,
+      "loss": 2.0652,
+      "step": 13917
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6720127939087868e-05,
+      "loss": 1.843,
+      "step": 13918
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.671968392655778e-05,
+      "loss": 1.7346,
+      "step": 13919
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6719239889872012e-05,
+      "loss": 1.8069,
+      "step": 13920
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.671879582903217e-05,
+      "loss": 2.2886,
+      "step": 13921
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.671835174403984e-05,
+      "loss": 2.1708,
+      "step": 13922
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.671790763489663e-05,
+      "loss": 1.8709,
+      "step": 13923
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.671746350160413e-05,
+      "loss": 1.8423,
+      "step": 13924
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6717019344163935e-05,
+      "loss": 1.7514,
+      "step": 13925
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6716575162577647e-05,
+      "loss": 1.801,
+      "step": 13926
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.671613095684686e-05,
+      "loss": 1.6548,
+      "step": 13927
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6715686726973167e-05,
+      "loss": 2.236,
+      "step": 13928
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.671524247295817e-05,
+      "loss": 1.9815,
+      "step": 13929
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6714798194803462e-05,
+      "loss": 1.7737,
+      "step": 13930
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6714353892510646e-05,
+      "loss": 2.0501,
+      "step": 13931
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6713909566081318e-05,
+      "loss": 1.7772,
+      "step": 13932
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6713465215517067e-05,
+      "loss": 1.8144,
+      "step": 13933
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.67130208408195e-05,
+      "loss": 1.7669,
+      "step": 13934
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6712576441990206e-05,
+      "loss": 1.8758,
+      "step": 13935
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.671213201903079e-05,
+      "loss": 1.938,
+      "step": 13936
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.671168757194285e-05,
+      "loss": 1.8217,
+      "step": 13937
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.671124310072798e-05,
+      "loss": 1.8685,
+      "step": 13938
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.6710798605387773e-05,
+      "loss": 2.1052,
+      "step": 13939
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6710354085923835e-05,
+      "loss": 2.2298,
+      "step": 13940
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.670990954233776e-05,
+      "loss": 2.0042,
+      "step": 13941
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.670946497463115e-05,
+      "loss": 2.0508,
+      "step": 13942
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6709020382805597e-05,
+      "loss": 2.0149,
+      "step": 13943
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6708575766862705e-05,
+      "loss": 1.7707,
+      "step": 13944
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.670813112680407e-05,
+      "loss": 1.6004,
+      "step": 13945
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6707686462631287e-05,
+      "loss": 1.7185,
+      "step": 13946
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.670724177434596e-05,
+      "loss": 2.3659,
+      "step": 13947
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.670679706194968e-05,
+      "loss": 1.7637,
+      "step": 13948
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6706352325444057e-05,
+      "loss": 2.1755,
+      "step": 13949
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.670590756483068e-05,
+      "loss": 1.7237,
+      "step": 13950
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.670546278011115e-05,
+      "loss": 1.8597,
+      "step": 13951
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6705017971287073e-05,
+      "loss": 1.8487,
+      "step": 13952
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.670457313836004e-05,
+      "loss": 1.5589,
+      "step": 13953
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6704128281331648e-05,
+      "loss": 2.1752,
+      "step": 13954
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6703683400203502e-05,
+      "loss": 1.7719,
+      "step": 13955
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.67032384949772e-05,
+      "loss": 1.8798,
+      "step": 13956
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.670279356565434e-05,
+      "loss": 1.9981,
+      "step": 13957
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6702348612236528e-05,
+      "loss": 1.7429,
+      "step": 13958
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.670190363472535e-05,
+      "loss": 1.6467,
+      "step": 13959
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6701458633122418e-05,
+      "loss": 1.986,
+      "step": 13960
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.6701013607429323e-05,
+      "loss": 1.8792,
+      "step": 13961
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6700568557647674e-05,
+      "loss": 1.8595,
+      "step": 13962
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6700123483779065e-05,
+      "loss": 1.9148,
+      "step": 13963
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6699678385825095e-05,
+      "loss": 1.9007,
+      "step": 13964
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6699233263787366e-05,
+      "loss": 2.0077,
+      "step": 13965
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.6698788117667476e-05,
+      "loss": 2.083,
+      "step": 13966
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6698342947467027e-05,
+      "loss": 2.0664,
+      "step": 13967
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6697897753187622e-05,
+      "loss": 1.9998,
+      "step": 13968
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6697452534830856e-05,
+      "loss": 1.973,
+      "step": 13969
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6697007292398334e-05,
+      "loss": 1.9014,
+      "step": 13970
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6696562025891654e-05,
+      "loss": 1.8507,
+      "step": 13971
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6696116735312417e-05,
+      "loss": 1.7886,
+      "step": 13972
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6695671420662224e-05,
+      "loss": 1.8619,
+      "step": 13973
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6695226081942672e-05,
+      "loss": 1.8282,
+      "step": 13974
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6694780719155374e-05,
+      "loss": 2.1743,
+      "step": 13975
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6694335332301914e-05,
+      "loss": 1.9433,
+      "step": 13976
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6693889921383906e-05,
+      "loss": 2.1584,
+      "step": 13977
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6693444486402946e-05,
+      "loss": 1.7797,
+      "step": 13978
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.669299902736064e-05,
+      "loss": 2.0399,
+      "step": 13979
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.669255354425858e-05,
+      "loss": 1.74,
+      "step": 13980
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.669210803709837e-05,
+      "loss": 1.8815,
+      "step": 13981
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.669166250588162e-05,
+      "loss": 1.9969,
+      "step": 13982
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6691216950609926e-05,
+      "loss": 1.9805,
+      "step": 13983
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.669077137128489e-05,
+      "loss": 1.8814,
+      "step": 13984
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.669032576790811e-05,
+      "loss": 1.8132,
+      "step": 13985
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.6689880140481193e-05,
+      "loss": 2.3745,
+      "step": 13986
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6689434489005737e-05,
+      "loss": 2.0793,
+      "step": 13987
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.6688988813483344e-05,
+      "loss": 2.3625,
+      "step": 13988
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6688543113915624e-05,
+      "loss": 1.899,
+      "step": 13989
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.6688097390304168e-05,
+      "loss": 1.859,
+      "step": 13990
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6687651642650587e-05,
+      "loss": 1.786,
+      "step": 13991
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6687205870956484e-05,
+      "loss": 1.987,
+      "step": 13992
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.668676007522345e-05,
+      "loss": 1.9359,
+      "step": 13993
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.6686314255453097e-05,
+      "loss": 2.2607,
+      "step": 13994
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6685868411647026e-05,
+      "loss": 1.8193,
+      "step": 13995
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6685422543806836e-05,
+      "loss": 1.8242,
+      "step": 13996
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6684976651934135e-05,
+      "loss": 2.1988,
+      "step": 13997
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6684530736030528e-05,
+      "loss": 1.9585,
+      "step": 13998
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.668408479609761e-05,
+      "loss": 1.9842,
+      "step": 13999
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.668363883213699e-05,
+      "loss": 1.8869,
+      "step": 14000
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6683192844150265e-05,
+      "loss": 1.6507,
+      "step": 14001
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.668274683213905e-05,
+      "loss": 1.665,
+      "step": 14002
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6682300796104933e-05,
+      "loss": 1.7445,
+      "step": 14003
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6681854736049528e-05,
+      "loss": 1.5982,
+      "step": 14004
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6681408651974437e-05,
+      "loss": 2.4351,
+      "step": 14005
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.668096254388126e-05,
+      "loss": 1.5482,
+      "step": 14006
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.66805164117716e-05,
+      "loss": 1.9924,
+      "step": 14007
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6680070255647067e-05,
+      "loss": 1.672,
+      "step": 14008
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.667962407550926e-05,
+      "loss": 2.1356,
+      "step": 14009
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6679177871359783e-05,
+      "loss": 2.0751,
+      "step": 14010
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6678731643200247e-05,
+      "loss": 1.7253,
+      "step": 14011
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6678285391032243e-05,
+      "loss": 1.7468,
+      "step": 14012
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6677839114857385e-05,
+      "loss": 1.8339,
+      "step": 14013
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6677392814677277e-05,
+      "loss": 2.1157,
+      "step": 14014
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.667694649049352e-05,
+      "loss": 1.6731,
+      "step": 14015
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6676500142307714e-05,
+      "loss": 1.9235,
+      "step": 14016
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6676053770121474e-05,
+      "loss": 2.327,
+      "step": 14017
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.66756073739364e-05,
+      "loss": 2.0021,
+      "step": 14018
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6675160953754094e-05,
+      "loss": 2.2106,
+      "step": 14019
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6674714509576163e-05,
+      "loss": 1.879,
+      "step": 14020
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6674268041404214e-05,
+      "loss": 2.1454,
+      "step": 14021
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.667382154923985e-05,
+      "loss": 2.001,
+      "step": 14022
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6673375033084674e-05,
+      "loss": 1.6441,
+      "step": 14023
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.667292849294029e-05,
+      "loss": 1.7011,
+      "step": 14024
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.667248192880831e-05,
+      "loss": 1.9262,
+      "step": 14025
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6672035340690337e-05,
+      "loss": 2.0507,
+      "step": 14026
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.667158872858797e-05,
+      "loss": 2.025,
+      "step": 14027
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6671142092502823e-05,
+      "loss": 1.9521,
+      "step": 14028
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.66706954324365e-05,
+      "loss": 1.9288,
+      "step": 14029
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.66702487483906e-05,
+      "loss": 1.5803,
+      "step": 14030
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6669802040366737e-05,
+      "loss": 1.8967,
+      "step": 14031
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.666935530836651e-05,
+      "loss": 1.941,
+      "step": 14032
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.666890855239153e-05,
+      "loss": 1.9023,
+      "step": 14033
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6668461772443397e-05,
+      "loss": 1.7849,
+      "step": 14034
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6668014968523722e-05,
+      "loss": 1.8622,
+      "step": 14035
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6667568140634115e-05,
+      "loss": 2.5891,
+      "step": 14036
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.666712128877617e-05,
+      "loss": 1.7254,
+      "step": 14037
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6666674412951506e-05,
+      "loss": 1.9072,
+      "step": 14038
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6666227513161725e-05,
+      "loss": 1.6485,
+      "step": 14039
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.666578058940843e-05,
+      "loss": 1.6043,
+      "step": 14040
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.666533364169323e-05,
+      "loss": 1.9318,
+      "step": 14041
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6664886670017734e-05,
+      "loss": 2.0097,
+      "step": 14042
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6664439674383546e-05,
+      "loss": 1.8172,
+      "step": 14043
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.666399265479227e-05,
+      "loss": 1.8295,
+      "step": 14044
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.666354561124552e-05,
+      "loss": 2.0108,
+      "step": 14045
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.66630985437449e-05,
+      "loss": 1.6668,
+      "step": 14046
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6662651452292014e-05,
+      "loss": 2.0771,
+      "step": 14047
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.666220433688847e-05,
+      "loss": 1.8824,
+      "step": 14048
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6661757197535885e-05,
+      "loss": 1.9623,
+      "step": 14049
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6661310034235852e-05,
+      "loss": 2.3266,
+      "step": 14050
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6660862846989983e-05,
+      "loss": 1.8199,
+      "step": 14051
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6660415635799893e-05,
+      "loss": 1.8144,
+      "step": 14052
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.665996840066718e-05,
+      "loss": 1.8429,
+      "step": 14053
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6659521141593456e-05,
+      "loss": 1.7643,
+      "step": 14054
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6659073858580328e-05,
+      "loss": 2.0076,
+      "step": 14055
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6658626551629405e-05,
+      "loss": 1.8998,
+      "step": 14056
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6658179220742297e-05,
+      "loss": 2.0031,
+      "step": 14057
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.6657731865920606e-05,
+      "loss": 1.8284,
+      "step": 14058
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6657284487165945e-05,
+      "loss": 1.8316,
+      "step": 14059
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.665683708447992e-05,
+      "loss": 1.7637,
+      "step": 14060
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.665638965786414e-05,
+      "loss": 1.7194,
+      "step": 14061
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6655942207320214e-05,
+      "loss": 1.6922,
+      "step": 14062
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.665549473284975e-05,
+      "loss": 1.6579,
+      "step": 14063
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6655047234454353e-05,
+      "loss": 2.1924,
+      "step": 14064
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6654599712135637e-05,
+      "loss": 1.9071,
+      "step": 14065
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.665415216589521e-05,
+      "loss": 1.7491,
+      "step": 14066
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.665370459573468e-05,
+      "loss": 1.9808,
+      "step": 14067
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.6653257001655652e-05,
+      "loss": 1.6938,
+      "step": 14068
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6652809383659744e-05,
+      "loss": 2.1999,
+      "step": 14069
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6652361741748558e-05,
+      "loss": 1.8832,
+      "step": 14070
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6651914075923704e-05,
+      "loss": 2.1072,
+      "step": 14071
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6651466386186788e-05,
+      "loss": 1.9,
+      "step": 14072
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.665101867253943e-05,
+      "loss": 1.7434,
+      "step": 14073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6650570934983232e-05,
+      "loss": 2.0263,
+      "step": 14074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.66501231735198e-05,
+      "loss": 1.8348,
+      "step": 14075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.6649675388150753e-05,
+      "loss": 1.9667,
+      "step": 14076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.664922757887769e-05,
+      "loss": 1.8504,
+      "step": 14077
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6648779745702228e-05,
+      "loss": 1.7454,
+      "step": 14078
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6648331888625978e-05,
+      "loss": 1.9993,
+      "step": 14079
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6647884007650544e-05,
+      "loss": 1.9624,
+      "step": 14080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6647436102777542e-05,
+      "loss": 1.4969,
+      "step": 14081
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6646988174008576e-05,
+      "loss": 2.0351,
+      "step": 14082
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.664654022134526e-05,
+      "loss": 1.4053,
+      "step": 14083
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6646092244789206e-05,
+      "loss": 1.6145,
+      "step": 14084
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.664564424434202e-05,
+      "loss": 1.933,
+      "step": 14085
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.664519622000532e-05,
+      "loss": 1.9204,
+      "step": 14086
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.6644748171780706e-05,
+      "loss": 1.8463,
+      "step": 14087
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6644300099669793e-05,
+      "loss": 2.0219,
+      "step": 14088
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6643852003674196e-05,
+      "loss": 1.8804,
+      "step": 14089
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6643403883795518e-05,
+      "loss": 1.9186,
+      "step": 14090
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.664295574003538e-05,
+      "loss": 1.7459,
+      "step": 14091
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.664250757239538e-05,
+      "loss": 1.8788,
+      "step": 14092
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6642059380877142e-05,
+      "loss": 1.8472,
+      "step": 14093
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.664161116548227e-05,
+      "loss": 1.9507,
+      "step": 14094
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6641162926212376e-05,
+      "loss": 1.6978,
+      "step": 14095
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.664071466306907e-05,
+      "loss": 2.3127,
+      "step": 14096
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.664026637605397e-05,
+      "loss": 1.7755,
+      "step": 14097
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6639818065168677e-05,
+      "loss": 1.7487,
+      "step": 14098
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.663936973041481e-05,
+      "loss": 1.9119,
+      "step": 14099
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.663892137179398e-05,
+      "loss": 1.765,
+      "step": 14100
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6638472989307798e-05,
+      "loss": 1.8049,
+      "step": 14101
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6638024582957878e-05,
+      "loss": 1.9604,
+      "step": 14102
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6637576152745823e-05,
+      "loss": 1.8726,
+      "step": 14103
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.6637127698673257e-05,
+      "loss": 1.922,
+      "step": 14104
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6636679220741782e-05,
+      "loss": 2.1072,
+      "step": 14105
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.663623071895302e-05,
+      "loss": 1.7561,
+      "step": 14106
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6635782193308573e-05,
+      "loss": 1.8451,
+      "step": 14107
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.663533364381006e-05,
+      "loss": 1.9412,
+      "step": 14108
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6634885070459095e-05,
+      "loss": 1.7173,
+      "step": 14109
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6634436473257287e-05,
+      "loss": 1.8236,
+      "step": 14110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6633987852206246e-05,
+      "loss": 1.8761,
+      "step": 14111
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6633539207307586e-05,
+      "loss": 1.9941,
+      "step": 14112
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6633090538562926e-05,
+      "loss": 1.3602,
+      "step": 14113
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.663264184597387e-05,
+      "loss": 1.7709,
+      "step": 14114
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6632193129542038e-05,
+      "loss": 1.947,
+      "step": 14115
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.663174438926904e-05,
+      "loss": 1.9148,
+      "step": 14116
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.663129562515649e-05,
+      "loss": 1.6033,
+      "step": 14117
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6630846837205995e-05,
+      "loss": 1.7898,
+      "step": 14118
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6630398025419178e-05,
+      "loss": 2.1061,
+      "step": 14119
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.662994918979765e-05,
+      "loss": 2.461,
+      "step": 14120
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.662950033034302e-05,
+      "loss": 1.992,
+      "step": 14121
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6629051447056904e-05,
+      "loss": 2.1383,
+      "step": 14122
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6628602539940916e-05,
+      "loss": 1.9509,
+      "step": 14123
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.662815360899667e-05,
+      "loss": 1.911,
+      "step": 14124
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.662770465422578e-05,
+      "loss": 2.0363,
+      "step": 14125
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.662725567562986e-05,
+      "loss": 1.4901,
+      "step": 14126
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.662680667321052e-05,
+      "loss": 1.821,
+      "step": 14127
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.662635764696938e-05,
+      "loss": 1.951,
+      "step": 14128
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6625908596908046e-05,
+      "loss": 2.1187,
+      "step": 14129
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6625459523028144e-05,
+      "loss": 1.6462,
+      "step": 14130
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.662501042533128e-05,
+      "loss": 2.1694,
+      "step": 14131
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.662456130381907e-05,
+      "loss": 1.8179,
+      "step": 14132
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.6624112158493127e-05,
+      "loss": 2.2442,
+      "step": 14133
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.662366298935507e-05,
+      "loss": 1.7192,
+      "step": 14134
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.662321379640651e-05,
+      "loss": 2.2095,
+      "step": 14135
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6622764579649062e-05,
+      "loss": 2.0911,
+      "step": 14136
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6622315339084344e-05,
+      "loss": 2.3711,
+      "step": 14137
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6621866074713965e-05,
+      "loss": 1.6681,
+      "step": 14138
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6621416786539545e-05,
+      "loss": 2.0081,
+      "step": 14139
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.66209674745627e-05,
+      "loss": 1.9329,
+      "step": 14140
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6620518138785038e-05,
+      "loss": 2.1614,
+      "step": 14141
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6620068779208183e-05,
+      "loss": 1.6664,
+      "step": 14142
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6619619395833744e-05,
+      "loss": 2.1193,
+      "step": 14143
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.661916998866334e-05,
+      "loss": 1.604,
+      "step": 14144
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6618720557698583e-05,
+      "loss": 1.9244,
+      "step": 14145
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6618271102941094e-05,
+      "loss": 1.6443,
+      "step": 14146
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6617821624392485e-05,
+      "loss": 1.821,
+      "step": 14147
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6617372122054373e-05,
+      "loss": 1.9652,
+      "step": 14148
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.661692259592837e-05,
+      "loss": 1.9366,
+      "step": 14149
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6616473046016097e-05,
+      "loss": 1.8729,
+      "step": 14150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.661602347231917e-05,
+      "loss": 1.8902,
+      "step": 14151
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.66155738748392e-05,
+      "loss": 1.9533,
+      "step": 14152
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6615124253577807e-05,
+      "loss": 1.7225,
+      "step": 14153
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.661467460853661e-05,
+      "loss": 1.9162,
+      "step": 14154
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6614224939717217e-05,
+      "loss": 1.8866,
+      "step": 14155
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6613775247121254e-05,
+      "loss": 1.8822,
+      "step": 14156
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.661332553075033e-05,
+      "loss": 2.0983,
+      "step": 14157
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6612875790606065e-05,
+      "loss": 2.0081,
+      "step": 14158
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6612426026690077e-05,
+      "loss": 1.7682,
+      "step": 14159
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.661197623900398e-05,
+      "loss": 1.895,
+      "step": 14160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6611526427549397e-05,
+      "loss": 2.1676,
+      "step": 14161
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6611076592327933e-05,
+      "loss": 1.9924,
+      "step": 14162
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6610626733341218e-05,
+      "loss": 1.9206,
+      "step": 14163
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.661017685059086e-05,
+      "loss": 1.9545,
+      "step": 14164
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.660972694407848e-05,
+      "loss": 1.7288,
+      "step": 14165
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6609277013805693e-05,
+      "loss": 1.7386,
+      "step": 14166
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.660882705977412e-05,
+      "loss": 1.8881,
+      "step": 14167
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6608377081985375e-05,
+      "loss": 1.7485,
+      "step": 14168
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6607927080441077e-05,
+      "loss": 1.8737,
+      "step": 14169
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.6607477055142848e-05,
+      "loss": 2.0025,
+      "step": 14170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.66070270060923e-05,
+      "loss": 1.9637,
+      "step": 14171
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.660657693329105e-05,
+      "loss": 1.7066,
+      "step": 14172
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6606126836740722e-05,
+      "loss": 2.0108,
+      "step": 14173
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6605676716442927e-05,
+      "loss": 1.5753,
+      "step": 14174
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6605226572399288e-05,
+      "loss": 1.9097,
+      "step": 14175
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.660477640461142e-05,
+      "loss": 1.9731,
+      "step": 14176
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6604326213080947e-05,
+      "loss": 1.9225,
+      "step": 14177
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6603875997809478e-05,
+      "loss": 1.9373,
+      "step": 14178
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.660342575879864e-05,
+      "loss": 2.1689,
+      "step": 14179
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6602975496050048e-05,
+      "loss": 1.8625,
+      "step": 14180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6602525209565317e-05,
+      "loss": 1.7466,
+      "step": 14181
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6602074899346072e-05,
+      "loss": 1.8849,
+      "step": 14182
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.660162456539393e-05,
+      "loss": 2.0908,
+      "step": 14183
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6601174207710505e-05,
+      "loss": 2.2196,
+      "step": 14184
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6600723826297423e-05,
+      "loss": 1.6739,
+      "step": 14185
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.66002734211563e-05,
+      "loss": 1.9477,
+      "step": 14186
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6599822992288754e-05,
+      "loss": 2.1367,
+      "step": 14187
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6599372539696406e-05,
+      "loss": 1.7053,
+      "step": 14188
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6598922063380875e-05,
+      "loss": 1.7824,
+      "step": 14189
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6598471563343783e-05,
+      "loss": 1.9681,
+      "step": 14190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.659802103958674e-05,
+      "loss": 1.6973,
+      "step": 14191
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.659757049211137e-05,
+      "loss": 1.8254,
+      "step": 14192
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.65971199209193e-05,
+      "loss": 1.9805,
+      "step": 14193
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.659666932601214e-05,
+      "loss": 2.0389,
+      "step": 14194
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6596218707391517e-05,
+      "loss": 1.8921,
+      "step": 14195
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.659576806505905e-05,
+      "loss": 1.9677,
+      "step": 14196
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.659531739901635e-05,
+      "loss": 1.8271,
+      "step": 14197
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.659486670926505e-05,
+      "loss": 1.7673,
+      "step": 14198
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6594415995806762e-05,
+      "loss": 1.7305,
+      "step": 14199
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6593965258643105e-05,
+      "loss": 1.8094,
+      "step": 14200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6593514497775704e-05,
+      "loss": 2.0783,
+      "step": 14201
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6593063713206178e-05,
+      "loss": 1.8006,
+      "step": 14202
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6592612904936147e-05,
+      "loss": 1.8342,
+      "step": 14203
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6592162072967233e-05,
+      "loss": 1.9957,
+      "step": 14204
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6591711217301054e-05,
+      "loss": 1.6874,
+      "step": 14205
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.659126033793923e-05,
+      "loss": 1.5989,
+      "step": 14206
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6590809434883386e-05,
+      "loss": 1.725,
+      "step": 14207
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.659035850813514e-05,
+      "loss": 1.773,
+      "step": 14208
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6589907557696116e-05,
+      "loss": 1.7445,
+      "step": 14209
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6589456583567934e-05,
+      "loss": 1.7358,
+      "step": 14210
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.658900558575221e-05,
+      "loss": 2.1384,
+      "step": 14211
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.658855456425057e-05,
+      "loss": 1.62,
+      "step": 14212
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6588103519064636e-05,
+      "loss": 1.7011,
+      "step": 14213
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.658765245019603e-05,
+      "loss": 2.0585,
+      "step": 14214
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.658720135764637e-05,
+      "loss": 1.8931,
+      "step": 14215
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6586750241417273e-05,
+      "loss": 1.8273,
+      "step": 14216
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.658629910151037e-05,
+      "loss": 1.821,
+      "step": 14217
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6585847937927282e-05,
+      "loss": 1.8267,
+      "step": 14218
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6585396750669625e-05,
+      "loss": 2.2177,
+      "step": 14219
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6584945539739027e-05,
+      "loss": 1.8325,
+      "step": 14220
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6584494305137108e-05,
+      "loss": 2.1558,
+      "step": 14221
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6584043046865486e-05,
+      "loss": 1.6415,
+      "step": 14222
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6583591764925785e-05,
+      "loss": 1.9904,
+      "step": 14223
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.658314045931963e-05,
+      "loss": 1.6992,
+      "step": 14224
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6582689130048644e-05,
+      "loss": 1.9711,
+      "step": 14225
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.6582237777114446e-05,
+      "loss": 2.3524,
+      "step": 14226
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.658178640051866e-05,
+      "loss": 1.7147,
+      "step": 14227
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6581335000262904e-05,
+      "loss": 1.951,
+      "step": 14228
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6580883576348813e-05,
+      "loss": 1.4816,
+      "step": 14229
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6580432128778e-05,
+      "loss": 1.8879,
+      "step": 14230
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6579980657552086e-05,
+      "loss": 2.2146,
+      "step": 14231
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.65795291626727e-05,
+      "loss": 2.2329,
+      "step": 14232
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.657907764414146e-05,
+      "loss": 1.6834,
+      "step": 14233
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6578626101959997e-05,
+      "loss": 2.2642,
+      "step": 14234
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6578174536129926e-05,
+      "loss": 1.9992,
+      "step": 14235
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.657772294665287e-05,
+      "loss": 2.0765,
+      "step": 14236
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.6577271333530463e-05,
+      "loss": 1.7229,
+      "step": 14237
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6576819696764314e-05,
+      "loss": 1.9171,
+      "step": 14238
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.657636803635606e-05,
+      "loss": 2.0881,
+      "step": 14239
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6575916352307314e-05,
+      "loss": 1.9494,
+      "step": 14240
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6575464644619702e-05,
+      "loss": 2.1055,
+      "step": 14241
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6575012913294853e-05,
+      "loss": 1.7895,
+      "step": 14242
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6574561158334387e-05,
+      "loss": 1.5188,
+      "step": 14243
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.657410937973993e-05,
+      "loss": 1.897,
+      "step": 14244
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6573657577513104e-05,
+      "loss": 2.0922,
+      "step": 14245
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.657320575165553e-05,
+      "loss": 1.7862,
+      "step": 14246
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6572753902168836e-05,
+      "loss": 2.8533,
+      "step": 14247
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.657230202905465e-05,
+      "loss": 1.9425,
+      "step": 14248
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.657185013231459e-05,
+      "loss": 1.9181,
+      "step": 14249
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6571398211950286e-05,
+      "loss": 2.2151,
+      "step": 14250
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6570946267963355e-05,
+      "loss": 1.7963,
+      "step": 14251
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6570494300355427e-05,
+      "loss": 1.8026,
+      "step": 14252
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6570042309128127e-05,
+      "loss": 1.8726,
+      "step": 14253
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6569590294283082e-05,
+      "loss": 1.9607,
+      "step": 14254
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.656913825582191e-05,
+      "loss": 1.843,
+      "step": 14255
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6568686193746238e-05,
+      "loss": 2.0206,
+      "step": 14256
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6568234108057695e-05,
+      "loss": 1.7428,
+      "step": 14257
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6567781998757906e-05,
+      "loss": 1.8032,
+      "step": 14258
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.656732986584849e-05,
+      "loss": 2.4231,
+      "step": 14259
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6566877709331078e-05,
+      "loss": 2.0985,
+      "step": 14260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6566425529207293e-05,
+      "loss": 1.7147,
+      "step": 14261
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6565973325478762e-05,
+      "loss": 1.9063,
+      "step": 14262
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6565521098147107e-05,
+      "loss": 1.8782,
+      "step": 14263
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6565068847213962e-05,
+      "loss": 2.1378,
+      "step": 14264
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6564616572680942e-05,
+      "loss": 2.0606,
+      "step": 14265
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6564164274549678e-05,
+      "loss": 2.0648,
+      "step": 14266
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6563711952821798e-05,
+      "loss": 1.728,
+      "step": 14267
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6563259607498924e-05,
+      "loss": 2.081,
+      "step": 14268
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.6562807238582685e-05,
+      "loss": 1.6502,
+      "step": 14269
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.656235484607471e-05,
+      "loss": 1.8641,
+      "step": 14270
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6561902429976615e-05,
+      "loss": 1.8052,
+      "step": 14271
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6561449990290035e-05,
+      "loss": 1.9144,
+      "step": 14272
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6560997527016593e-05,
+      "loss": 1.9454,
+      "step": 14273
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6560545040157918e-05,
+      "loss": 2.0398,
+      "step": 14274
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6560092529715634e-05,
+      "loss": 1.7873,
+      "step": 14275
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6559639995691366e-05,
+      "loss": 1.6747,
+      "step": 14276
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.655918743808675e-05,
+      "loss": 1.5588,
+      "step": 14277
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.6558734856903406e-05,
+      "loss": 1.6482,
+      "step": 14278
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6558282252142956e-05,
+      "loss": 1.7349,
+      "step": 14279
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6557829623807035e-05,
+      "loss": 1.8236,
+      "step": 14280
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.655737697189727e-05,
+      "loss": 1.9025,
+      "step": 14281
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.655692429641528e-05,
+      "loss": 2.1769,
+      "step": 14282
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.65564715973627e-05,
+      "loss": 1.8237,
+      "step": 14283
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6556018874741155e-05,
+      "loss": 1.9641,
+      "step": 14284
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6555566128552274e-05,
+      "loss": 1.8486,
+      "step": 14285
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6555113358797684e-05,
+      "loss": 2.0797,
+      "step": 14286
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.655466056547901e-05,
+      "loss": 1.8019,
+      "step": 14287
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.655420774859788e-05,
+      "loss": 1.9485,
+      "step": 14288
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6553754908155927e-05,
+      "loss": 2.0688,
+      "step": 14289
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.655330204415477e-05,
+      "loss": 1.8774,
+      "step": 14290
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6552849156596045e-05,
+      "loss": 1.9031,
+      "step": 14291
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6552396245481376e-05,
+      "loss": 2.1037,
+      "step": 14292
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.655194331081239e-05,
+      "loss": 1.9525,
+      "step": 14293
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6551490352590724e-05,
+      "loss": 1.6752,
+      "step": 14294
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6551037370817996e-05,
+      "loss": 1.9464,
+      "step": 14295
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6550584365495835e-05,
+      "loss": 2.0236,
+      "step": 14296
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6550131336625876e-05,
+      "loss": 1.6527,
+      "step": 14297
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6549678284209744e-05,
+      "loss": 1.8821,
+      "step": 14298
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.6549225208249066e-05,
+      "loss": 1.9054,
+      "step": 14299
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.654877210874547e-05,
+      "loss": 1.8361,
+      "step": 14300
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6548318985700594e-05,
+      "loss": 1.8604,
+      "step": 14301
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6547865839116054e-05,
+      "loss": 1.9628,
+      "step": 14302
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6547412668993486e-05,
+      "loss": 1.8662,
+      "step": 14303
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6546959475334518e-05,
+      "loss": 2.0392,
+      "step": 14304
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.654650625814078e-05,
+      "loss": 1.8511,
+      "step": 14305
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.65460530174139e-05,
+      "loss": 1.8374,
+      "step": 14306
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6545599753155507e-05,
+      "loss": 1.888,
+      "step": 14307
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.654514646536723e-05,
+      "loss": 1.7071,
+      "step": 14308
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.65446931540507e-05,
+      "loss": 1.7468,
+      "step": 14309
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6544239819207546e-05,
+      "loss": 1.9078,
+      "step": 14310
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6543786460839396e-05,
+      "loss": 1.9148,
+      "step": 14311
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6543333078947884e-05,
+      "loss": 1.7568,
+      "step": 14312
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6542879673534636e-05,
+      "loss": 2.1837,
+      "step": 14313
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6542426244601283e-05,
+      "loss": 1.5483,
+      "step": 14314
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6541972792149453e-05,
+      "loss": 1.6776,
+      "step": 14315
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.654151931618078e-05,
+      "loss": 1.9361,
+      "step": 14316
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.654106581669689e-05,
+      "loss": 2.2035,
+      "step": 14317
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.654061229369942e-05,
+      "loss": 2.0343,
+      "step": 14318
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6540158747189992e-05,
+      "loss": 1.9741,
+      "step": 14319
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.653970517717024e-05,
+      "loss": 1.9396,
+      "step": 14320
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6539251583641796e-05,
+      "loss": 1.8382,
+      "step": 14321
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6538797966606288e-05,
+      "loss": 2.1209,
+      "step": 14322
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6538344326065344e-05,
+      "loss": 1.5984,
+      "step": 14323
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6537890662020605e-05,
+      "loss": 1.858,
+      "step": 14324
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6537436974473693e-05,
+      "loss": 1.5869,
+      "step": 14325
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.653698326342624e-05,
+      "loss": 1.9049,
+      "step": 14326
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.653652952887988e-05,
+      "loss": 1.9019,
+      "step": 14327
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6536075770836243e-05,
+      "loss": 1.912,
+      "step": 14328
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6535621989296954e-05,
+      "loss": 1.7889,
+      "step": 14329
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6535168184263652e-05,
+      "loss": 2.0066,
+      "step": 14330
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.653471435573797e-05,
+      "loss": 1.6986,
+      "step": 14331
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6534260503721528e-05,
+      "loss": 1.8549,
+      "step": 14332
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6533806628215972e-05,
+      "loss": 2.0685,
+      "step": 14333
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.653335272922292e-05,
+      "loss": 1.7167,
+      "step": 14334
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6532898806744015e-05,
+      "loss": 1.8459,
+      "step": 14335
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.653244486078088e-05,
+      "loss": 1.5616,
+      "step": 14336
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6531990891335152e-05,
+      "loss": 1.9988,
+      "step": 14337
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6531536898408465e-05,
+      "loss": 2.2462,
+      "step": 14338
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6531082882002444e-05,
+      "loss": 2.2078,
+      "step": 14339
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6530628842118727e-05,
+      "loss": 2.0462,
+      "step": 14340
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.653017477875894e-05,
+      "loss": 2.2817,
+      "step": 14341
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.652972069192472e-05,
+      "loss": 1.6441,
+      "step": 14342
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.65292665816177e-05,
+      "loss": 1.8859,
+      "step": 14343
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.652881244783951e-05,
+      "loss": 1.902,
+      "step": 14344
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.6528358290591784e-05,
+      "loss": 1.9479,
+      "step": 14345
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.652790410987615e-05,
+      "loss": 1.9571,
+      "step": 14346
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.652744990569425e-05,
+      "loss": 1.6746,
+      "step": 14347
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6526995678047712e-05,
+      "loss": 1.905,
+      "step": 14348
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6526541426938165e-05,
+      "loss": 2.0692,
+      "step": 14349
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6526087152367246e-05,
+      "loss": 1.9964,
+      "step": 14350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6525632854336584e-05,
+      "loss": 1.9463,
+      "step": 14351
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6525178532847816e-05,
+      "loss": 2.1756,
+      "step": 14352
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6524724187902577e-05,
+      "loss": 1.9887,
+      "step": 14353
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.6524269819502495e-05,
+      "loss": 1.906,
+      "step": 14354
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.652381542764921e-05,
+      "loss": 1.9538,
+      "step": 14355
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6523361012344348e-05,
+      "loss": 1.621,
+      "step": 14356
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6522906573589547e-05,
+      "loss": 1.7895,
+      "step": 14357
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6522452111386437e-05,
+      "loss": 2.0352,
+      "step": 14358
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6521997625736656e-05,
+      "loss": 2.2343,
+      "step": 14359
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.652154311664184e-05,
+      "loss": 1.9889,
+      "step": 14360
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6521088584103614e-05,
+      "loss": 1.6396,
+      "step": 14361
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.652063402812362e-05,
+      "loss": 1.7014,
+      "step": 14362
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.652017944870348e-05,
+      "loss": 2.0565,
+      "step": 14363
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6519724845844844e-05,
+      "loss": 1.971,
+      "step": 14364
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.651927021954934e-05,
+      "loss": 1.8452,
+      "step": 14365
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.65188155698186e-05,
+      "loss": 1.9043,
+      "step": 14366
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.651836089665426e-05,
+      "loss": 1.879,
+      "step": 14367
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6517906200057953e-05,
+      "loss": 1.6198,
+      "step": 14368
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6517451480031317e-05,
+      "loss": 1.923,
+      "step": 14369
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6516996736575982e-05,
+      "loss": 1.5309,
+      "step": 14370
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6516541969693583e-05,
+      "loss": 1.7413,
+      "step": 14371
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6516087179385763e-05,
+      "loss": 1.8825,
+      "step": 14372
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6515632365654144e-05,
+      "loss": 1.8364,
+      "step": 14373
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.651517752850037e-05,
+      "loss": 2.3306,
+      "step": 14374
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6514722667926075e-05,
+      "loss": 1.681,
+      "step": 14375
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6514267783932894e-05,
+      "loss": 1.8172,
+      "step": 14376
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6513812876522454e-05,
+      "loss": 1.8817,
+      "step": 14377
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6513357945696405e-05,
+      "loss": 1.8841,
+      "step": 14378
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.651290299145637e-05,
+      "loss": 1.9794,
+      "step": 14379
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.651244801380399e-05,
+      "loss": 1.7454,
+      "step": 14380
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6511993012740898e-05,
+      "loss": 1.9357,
+      "step": 14381
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6511537988268736e-05,
+      "loss": 1.699,
+      "step": 14382
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.651108294038913e-05,
+      "loss": 1.702,
+      "step": 14383
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.651062786910372e-05,
+      "loss": 1.8955,
+      "step": 14384
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.651017277441415e-05,
+      "loss": 1.9714,
+      "step": 14385
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.650971765632204e-05,
+      "loss": 1.8535,
+      "step": 14386
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.650926251482904e-05,
+      "loss": 1.6305,
+      "step": 14387
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.650880734993678e-05,
+      "loss": 1.829,
+      "step": 14388
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.650835216164689e-05,
+      "loss": 1.7678,
+      "step": 14389
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.650789694996102e-05,
+      "loss": 1.9371,
+      "step": 14390
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6507441714880797e-05,
+      "loss": 1.7132,
+      "step": 14391
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6506986456407866e-05,
+      "loss": 1.9772,
+      "step": 14392
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.650653117454385e-05,
+      "loss": 1.8663,
+      "step": 14393
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6506075869290397e-05,
+      "loss": 1.912,
+      "step": 14394
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6505620540649138e-05,
+      "loss": 1.8413,
+      "step": 14395
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.650516518862171e-05,
+      "loss": 2.0234,
+      "step": 14396
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.650470981320976e-05,
+      "loss": 1.8938,
+      "step": 14397
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.650425441441491e-05,
+      "loss": 1.9609,
+      "step": 14398
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6503798992238803e-05,
+      "loss": 1.6168,
+      "step": 14399
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6503343546683076e-05,
+      "loss": 2.0358,
+      "step": 14400
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.650288807774937e-05,
+      "loss": 1.7706,
+      "step": 14401
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6502432585439318e-05,
+      "loss": 2.0552,
+      "step": 14402
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.650197706975456e-05,
+      "loss": 1.8364,
+      "step": 14403
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6501521530696728e-05,
+      "loss": 2.0269,
+      "step": 14404
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.650106596826747e-05,
+      "loss": 1.7012,
+      "step": 14405
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6500610382468413e-05,
+      "loss": 2.0448,
+      "step": 14406
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.65001547733012e-05,
+      "loss": 1.9653,
+      "step": 14407
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6499699140767468e-05,
+      "loss": 1.9481,
+      "step": 14408
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.6499243484868853e-05,
+      "loss": 1.6026,
+      "step": 14409
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6498787805606998e-05,
+      "loss": 1.8509,
+      "step": 14410
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.6498332102983535e-05,
+      "loss": 1.8834,
+      "step": 14411
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6497876377000105e-05,
+      "loss": 1.9096,
+      "step": 14412
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6497420627658346e-05,
+      "loss": 2.1293,
+      "step": 14413
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6496964854959898e-05,
+      "loss": 1.7776,
+      "step": 14414
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.64965090589064e-05,
+      "loss": 1.7981,
+      "step": 14415
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6496053239499485e-05,
+      "loss": 1.8757,
+      "step": 14416
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6495597396740793e-05,
+      "loss": 2.1179,
+      "step": 14417
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6495141530631968e-05,
+      "loss": 1.7262,
+      "step": 14418
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6494685641174646e-05,
+      "loss": 1.7171,
+      "step": 14419
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.649422972837046e-05,
+      "loss": 1.9872,
+      "step": 14420
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6493773792221056e-05,
+      "loss": 2.0637,
+      "step": 14421
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6493317832728075e-05,
+      "loss": 1.8076,
+      "step": 14422
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.649286184989315e-05,
+      "loss": 1.8881,
+      "step": 14423
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6492405843717923e-05,
+      "loss": 1.959,
+      "step": 14424
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6491949814204028e-05,
+      "loss": 1.9098,
+      "step": 14425
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6491493761353112e-05,
+      "loss": 2.0216,
+      "step": 14426
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.649103768516681e-05,
+      "loss": 1.8242,
+      "step": 14427
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6490581585646764e-05,
+      "loss": 1.783,
+      "step": 14428
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6490125462794613e-05,
+      "loss": 1.7773,
+      "step": 14429
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6489669316611992e-05,
+      "loss": 2.1454,
+      "step": 14430
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6489213147100546e-05,
+      "loss": 1.7581,
+      "step": 14431
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6488756954261914e-05,
+      "loss": 1.8905,
+      "step": 14432
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6488300738097737e-05,
+      "loss": 1.9858,
+      "step": 14433
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.648784449860965e-05,
+      "loss": 2.3168,
+      "step": 14434
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.64873882357993e-05,
+      "loss": 2.0521,
+      "step": 14435
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6486931949668324e-05,
+      "loss": 1.6546,
+      "step": 14436
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6486475640218354e-05,
+      "loss": 2.1669,
+      "step": 14437
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6486019307451045e-05,
+      "loss": 2.0647,
+      "step": 14438
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6485562951368033e-05,
+      "loss": 1.9239,
+      "step": 14439
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.648510657197095e-05,
+      "loss": 1.7964,
+      "step": 14440
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6484650169261443e-05,
+      "loss": 1.6408,
+      "step": 14441
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6484193743241153e-05,
+      "loss": 1.7406,
+      "step": 14442
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6483737293911722e-05,
+      "loss": 2.0325,
+      "step": 14443
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6483280821274788e-05,
+      "loss": 2.066,
+      "step": 14444
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6482824325331992e-05,
+      "loss": 1.7865,
+      "step": 14445
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.648236780608498e-05,
+      "loss": 1.8344,
+      "step": 14446
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.648191126353538e-05,
+      "loss": 1.7744,
+      "step": 14447
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6481454697684846e-05,
+      "loss": 1.7877,
+      "step": 14448
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.648099810853502e-05,
+      "loss": 1.8865,
+      "step": 14449
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.648054149608753e-05,
+      "loss": 1.7668,
+      "step": 14450
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.648008486034403e-05,
+      "loss": 1.6855,
+      "step": 14451
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.647962820130616e-05,
+      "loss": 1.9121,
+      "step": 14452
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6479171518975556e-05,
+      "loss": 1.854,
+      "step": 14453
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6478714813353864e-05,
+      "loss": 2.0172,
+      "step": 14454
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6478258084442723e-05,
+      "loss": 1.9404,
+      "step": 14455
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6477801332243777e-05,
+      "loss": 1.7698,
+      "step": 14456
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.647734455675867e-05,
+      "loss": 1.9771,
+      "step": 14457
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6476887757989035e-05,
+      "loss": 2.1795,
+      "step": 14458
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6476430935936525e-05,
+      "loss": 1.939,
+      "step": 14459
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6475974090602776e-05,
+      "loss": 2.0081,
+      "step": 14460
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.647551722198943e-05,
+      "loss": 1.7737,
+      "step": 14461
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6475060330098133e-05,
+      "loss": 2.248,
+      "step": 14462
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6474603414930526e-05,
+      "loss": 1.901,
+      "step": 14463
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.647414647648825e-05,
+      "loss": 1.9931,
+      "step": 14464
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6473689514772945e-05,
+      "loss": 1.958,
+      "step": 14465
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.647323252978626e-05,
+      "loss": 1.6834,
+      "step": 14466
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6472775521529838e-05,
+      "loss": 1.9382,
+      "step": 14467
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6472318490005318e-05,
+      "loss": 1.7944,
+      "step": 14468
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.647186143521434e-05,
+      "loss": 1.9633,
+      "step": 14469
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6471404357158554e-05,
+      "loss": 2.1231,
+      "step": 14470
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6470947255839597e-05,
+      "loss": 1.7571,
+      "step": 14471
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.647049013125912e-05,
+      "loss": 1.6394,
+      "step": 14472
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6470032983418754e-05,
+      "loss": 2.1891,
+      "step": 14473
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6469575812320155e-05,
+      "loss": 1.9121,
+      "step": 14474
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.646911861796496e-05,
+      "loss": 1.8475,
+      "step": 14475
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6468661400354812e-05,
+      "loss": 2.0574,
+      "step": 14476
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6468204159491358e-05,
+      "loss": 2.1182,
+      "step": 14477
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.646774689537624e-05,
+      "loss": 2.1194,
+      "step": 14478
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.64672896080111e-05,
+      "loss": 1.9335,
+      "step": 14479
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6466832297397578e-05,
+      "loss": 1.9874,
+      "step": 14480
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.646637496353733e-05,
+      "loss": 1.7419,
+      "step": 14481
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6465917606431993e-05,
+      "loss": 1.8968,
+      "step": 14482
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.646546022608321e-05,
+      "loss": 1.9632,
+      "step": 14483
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6465002822492625e-05,
+      "loss": 1.7368,
+      "step": 14484
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.6464545395661888e-05,
+      "loss": 1.9249,
+      "step": 14485
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6464087945592635e-05,
+      "loss": 1.8153,
+      "step": 14486
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6463630472286518e-05,
+      "loss": 1.9134,
+      "step": 14487
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6463172975745176e-05,
+      "loss": 1.5373,
+      "step": 14488
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6462715455970255e-05,
+      "loss": 2.0954,
+      "step": 14489
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6462257912963402e-05,
+      "loss": 1.9899,
+      "step": 14490
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6461800346726258e-05,
+      "loss": 1.8573,
+      "step": 14491
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.646134275726047e-05,
+      "loss": 1.9776,
+      "step": 14492
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6460885144567686e-05,
+      "loss": 2.2013,
+      "step": 14493
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.6460427508649546e-05,
+      "loss": 1.8097,
+      "step": 14494
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6459969849507696e-05,
+      "loss": 1.8226,
+      "step": 14495
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.625,
+      "learning_rate": 1.645951216714378e-05,
+      "loss": 1.9918,
+      "step": 14496
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.645905446155945e-05,
+      "loss": 2.0749,
+      "step": 14497
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6458596732756345e-05,
+      "loss": 1.8108,
+      "step": 14498
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6458138980736117e-05,
+      "loss": 2.5432,
+      "step": 14499
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6457681205500396e-05,
+      "loss": 2.1409,
+      "step": 14500
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.645722340705085e-05,
+      "loss": 1.8431,
+      "step": 14501
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.6456765585389104e-05,
+      "loss": 1.749,
+      "step": 14502
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6456307740516815e-05,
+      "loss": 1.6834,
+      "step": 14503
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.645584987243563e-05,
+      "loss": 2.3225,
+      "step": 14504
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6455391981147188e-05,
+      "loss": 1.9033,
+      "step": 14505
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6454934066653138e-05,
+      "loss": 1.7081,
+      "step": 14506
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.645447612895513e-05,
+      "loss": 1.7589,
+      "step": 14507
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6454018168054804e-05,
+      "loss": 1.6605,
+      "step": 14508
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.645356018395381e-05,
+      "loss": 2.1795,
+      "step": 14509
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6453102176653792e-05,
+      "loss": 1.8941,
+      "step": 14510
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.64526441461564e-05,
+      "loss": 1.981,
+      "step": 14511
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6452186092463275e-05,
+      "loss": 1.796,
+      "step": 14512
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6451728015576072e-05,
+      "loss": 1.7198,
+      "step": 14513
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.645126991549643e-05,
+      "loss": 2.027,
+      "step": 14514
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6450811792225997e-05,
+      "loss": 1.6835,
+      "step": 14515
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6450353645766422e-05,
+      "loss": 1.9137,
+      "step": 14516
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.644989547611935e-05,
+      "loss": 1.634,
+      "step": 14517
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.644943728328643e-05,
+      "loss": 1.7095,
+      "step": 14518
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6448979067269307e-05,
+      "loss": 2.128,
+      "step": 14519
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6448520828069633e-05,
+      "loss": 1.994,
+      "step": 14520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.644806256568905e-05,
+      "loss": 1.5716,
+      "step": 14521
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6447604280129207e-05,
+      "loss": 1.5165,
+      "step": 14522
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.644714597139175e-05,
+      "loss": 1.8799,
+      "step": 14523
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.644668763947833e-05,
+      "loss": 1.9075,
+      "step": 14524
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.75,
+      "learning_rate": 1.644622928439059e-05,
+      "loss": 1.8977,
+      "step": 14525
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6445770906130184e-05,
+      "loss": 2.2467,
+      "step": 14526
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6445312504698753e-05,
+      "loss": 1.7759,
+      "step": 14527
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6444854080097946e-05,
+      "loss": 1.6929,
+      "step": 14528
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.644439563232942e-05,
+      "loss": 1.6268,
+      "step": 14529
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.644393716139481e-05,
+      "loss": 1.8615,
+      "step": 14530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.644347866729577e-05,
+      "loss": 2.0092,
+      "step": 14531
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.644302015003395e-05,
+      "loss": 1.6735,
+      "step": 14532
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6442561609610995e-05,
+      "loss": 2.0777,
+      "step": 14533
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.644210304602855e-05,
+      "loss": 1.7334,
+      "step": 14534
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6441644459288277e-05,
+      "loss": 2.2357,
+      "step": 14535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.644118584939181e-05,
+      "loss": 1.7981,
+      "step": 14536
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6440727216340808e-05,
+      "loss": 2.1208,
+      "step": 14537
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.644026856013691e-05,
+      "loss": 1.314,
+      "step": 14538
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6439809880781768e-05,
+      "loss": 1.6204,
+      "step": 14539
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6439351178277037e-05,
+      "loss": 1.423,
+      "step": 14540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.643889245262436e-05,
+      "loss": 1.968,
+      "step": 14541
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.643843370382539e-05,
+      "loss": 1.76,
+      "step": 14542
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.643797493188177e-05,
+      "loss": 1.6558,
+      "step": 14543
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6437516136795154e-05,
+      "loss": 2.0474,
+      "step": 14544
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6437057318567185e-05,
+      "loss": 1.8417,
+      "step": 14545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6436598477199527e-05,
+      "loss": 1.8767,
+      "step": 14546
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.643613961269381e-05,
+      "loss": 1.839,
+      "step": 14547
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.64356807250517e-05,
+      "loss": 2.147,
+      "step": 14548
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6435221814274838e-05,
+      "loss": 1.6255,
+      "step": 14549
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6434762880364873e-05,
+      "loss": 1.8137,
+      "step": 14550
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6434303923323458e-05,
+      "loss": 1.6455,
+      "step": 14551
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6433844943152242e-05,
+      "loss": 1.8271,
+      "step": 14552
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.643338593985288e-05,
+      "loss": 1.7175,
+      "step": 14553
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6432926913427013e-05,
+      "loss": 1.746,
+      "step": 14554
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6432467863876294e-05,
+      "loss": 1.7847,
+      "step": 14555
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6432008791202373e-05,
+      "loss": 1.9374,
+      "step": 14556
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6431549695406905e-05,
+      "loss": 1.9414,
+      "step": 14557
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6431090576491537e-05,
+      "loss": 1.7525,
+      "step": 14558
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6430631434457916e-05,
+      "loss": 1.8037,
+      "step": 14559
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.64301722693077e-05,
+      "loss": 1.875,
+      "step": 14560
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6429713081042532e-05,
+      "loss": 1.6016,
+      "step": 14561
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.642925386966407e-05,
+      "loss": 1.9061,
+      "step": 14562
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6428794635173955e-05,
+      "loss": 1.701,
+      "step": 14563
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.642833537757385e-05,
+      "loss": 1.4732,
+      "step": 14564
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 1.8803,
+      "step": 14565
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.642741679305025e-05,
+      "loss": 2.0124,
+      "step": 14566
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6426957466130057e-05,
+      "loss": 1.8576,
+      "step": 14567
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6426498116106473e-05,
+      "loss": 1.9586,
+      "step": 14568
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.642603874298115e-05,
+      "loss": 1.7291,
+      "step": 14569
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6425579346755732e-05,
+      "loss": 1.5995,
+      "step": 14570
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.642511992743188e-05,
+      "loss": 2.168,
+      "step": 14571
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6424660485011245e-05,
+      "loss": 1.9942,
+      "step": 14572
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.642420101949547e-05,
+      "loss": 1.9774,
+      "step": 14573
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6423741530886213e-05,
+      "loss": 1.6784,
+      "step": 14574
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.6423282019185123e-05,
+      "loss": 1.8462,
+      "step": 14575
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6422822484393855e-05,
+      "loss": 1.765,
+      "step": 14576
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6422362926514057e-05,
+      "loss": 1.6027,
+      "step": 14577
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6421903345547387e-05,
+      "loss": 1.8993,
+      "step": 14578
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6421443741495487e-05,
+      "loss": 1.7289,
+      "step": 14579
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6420984114360023e-05,
+      "loss": 1.8877,
+      "step": 14580
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6420524464142636e-05,
+      "loss": 1.8062,
+      "step": 14581
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.642006479084498e-05,
+      "loss": 1.5892,
+      "step": 14582
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.6419605094468712e-05,
+      "loss": 2.3075,
+      "step": 14583
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.641914537501548e-05,
+      "loss": 1.7829,
+      "step": 14584
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.641868563248694e-05,
+      "loss": 1.8264,
+      "step": 14585
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6418225866884743e-05,
+      "loss": 1.9689,
+      "step": 14586
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.641776607821054e-05,
+      "loss": 1.7897,
+      "step": 14587
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6417306266465987e-05,
+      "loss": 1.9543,
+      "step": 14588
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.6416846431652734e-05,
+      "loss": 1.9837,
+      "step": 14589
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.641638657377244e-05,
+      "loss": 1.8426,
+      "step": 14590
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6415926692826748e-05,
+      "loss": 1.8152,
+      "step": 14591
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6415466788817324e-05,
+      "loss": 1.5635,
+      "step": 14592
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.641500686174581e-05,
+      "loss": 1.9961,
+      "step": 14593
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6414546911613866e-05,
+      "loss": 1.7612,
+      "step": 14594
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6414086938423138e-05,
+      "loss": 2.0115,
+      "step": 14595
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6413626942175285e-05,
+      "loss": 2.0607,
+      "step": 14596
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6413166922871965e-05,
+      "loss": 1.6967,
+      "step": 14597
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.641270688051482e-05,
+      "loss": 2.0832,
+      "step": 14598
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.6412246815105515e-05,
+      "loss": 2.2008,
+      "step": 14599
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.64117867266457e-05,
+      "loss": 2.0997,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6411326615137025e-05,
+      "loss": 2.2082,
+      "step": 14601
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6410866480581144e-05,
+      "loss": 1.9519,
+      "step": 14602
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6410406322979722e-05,
+      "loss": 1.7285,
+      "step": 14603
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.64099461423344e-05,
+      "loss": 1.8595,
+      "step": 14604
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6409485938646836e-05,
+      "loss": 1.8691,
+      "step": 14605
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.640902571191869e-05,
+      "loss": 1.5668,
+      "step": 14606
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.640856546215161e-05,
+      "loss": 1.6049,
+      "step": 14607
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6408105189347256e-05,
+      "loss": 1.8444,
+      "step": 14608
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6407644893507276e-05,
+      "loss": 1.9611,
+      "step": 14609
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.640718457463333e-05,
+      "loss": 1.6228,
+      "step": 14610
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6406724232727068e-05,
+      "loss": 1.6262,
+      "step": 14611
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6406263867790148e-05,
+      "loss": 2.2023,
+      "step": 14612
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.640580347982423e-05,
+      "loss": 2.0058,
+      "step": 14613
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6405343068830956e-05,
+      "loss": 1.8599,
+      "step": 14614
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.640488263481199e-05,
+      "loss": 1.8665,
+      "step": 14615
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.640442217776899e-05,
+      "loss": 1.8217,
+      "step": 14616
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6403961697703604e-05,
+      "loss": 1.708,
+      "step": 14617
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6403501194617487e-05,
+      "loss": 1.6457,
+      "step": 14618
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6403040668512305e-05,
+      "loss": 1.7152,
+      "step": 14619
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.64025801193897e-05,
+      "loss": 1.6686,
+      "step": 14620
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6402119547251337e-05,
+      "loss": 1.7289,
+      "step": 14621
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6401658952098864e-05,
+      "loss": 1.9489,
+      "step": 14622
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6401198333933943e-05,
+      "loss": 1.8372,
+      "step": 14623
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.640073769275823e-05,
+      "loss": 1.9466,
+      "step": 14624
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6400277028573376e-05,
+      "loss": 1.8271,
+      "step": 14625
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6399816341381043e-05,
+      "loss": 2.0859,
+      "step": 14626
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6399355631182882e-05,
+      "loss": 2.0834,
+      "step": 14627
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6398894897980553e-05,
+      "loss": 2.0907,
+      "step": 14628
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6398434141775706e-05,
+      "loss": 1.8295,
+      "step": 14629
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6397973362570004e-05,
+      "loss": 2.1932,
+      "step": 14630
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6397512560365098e-05,
+      "loss": 1.9918,
+      "step": 14631
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.639705173516265e-05,
+      "loss": 1.5517,
+      "step": 14632
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6396590886964314e-05,
+      "loss": 2.4541,
+      "step": 14633
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6396130015771746e-05,
+      "loss": 1.9792,
+      "step": 14634
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6395669121586604e-05,
+      "loss": 2.2307,
+      "step": 14635
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.639520820441054e-05,
+      "loss": 1.9463,
+      "step": 14636
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6394747264245222e-05,
+      "loss": 1.9085,
+      "step": 14637
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6394286301092296e-05,
+      "loss": 1.8128,
+      "step": 14638
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.639382531495342e-05,
+      "loss": 1.9525,
+      "step": 14639
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.639336430583026e-05,
+      "loss": 2.0865,
+      "step": 14640
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6392903273724462e-05,
+      "loss": 1.7319,
+      "step": 14641
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6392442218637693e-05,
+      "loss": 1.6003,
+      "step": 14642
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6391981140571602e-05,
+      "loss": 1.8814,
+      "step": 14643
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6391520039527855e-05,
+      "loss": 1.8172,
+      "step": 14644
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6391058915508102e-05,
+      "loss": 1.7524,
+      "step": 14645
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6390597768514003e-05,
+      "loss": 1.5249,
+      "step": 14646
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.639013659854722e-05,
+      "loss": 2.3533,
+      "step": 14647
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6389675405609403e-05,
+      "loss": 2.0047,
+      "step": 14648
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6389214189702214e-05,
+      "loss": 1.8679,
+      "step": 14649
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6388752950827314e-05,
+      "loss": 2.056,
+      "step": 14650
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6388291688986356e-05,
+      "loss": 1.8729,
+      "step": 14651
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6387830404181e-05,
+      "loss": 2.0515,
+      "step": 14652
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6387369096412905e-05,
+      "loss": 1.712,
+      "step": 14653
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.638690776568373e-05,
+      "loss": 1.8005,
+      "step": 14654
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.638644641199513e-05,
+      "loss": 2.2462,
+      "step": 14655
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6385985035348766e-05,
+      "loss": 1.6439,
+      "step": 14656
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6385523635746298e-05,
+      "loss": 1.8291,
+      "step": 14657
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.638506221318938e-05,
+      "loss": 1.5066,
+      "step": 14658
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6384600767679674e-05,
+      "loss": 1.3939,
+      "step": 14659
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.638413929921884e-05,
+      "loss": 2.0901,
+      "step": 14660
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6383677807808533e-05,
+      "loss": 1.7912,
+      "step": 14661
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6383216293450414e-05,
+      "loss": 2.0829,
+      "step": 14662
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.6382754756146142e-05,
+      "loss": 1.9682,
+      "step": 14663
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6382293195897377e-05,
+      "loss": 1.8991,
+      "step": 14664
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6381831612705776e-05,
+      "loss": 1.5949,
+      "step": 14665
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6381370006573e-05,
+      "loss": 1.8543,
+      "step": 14666
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6380908377500714e-05,
+      "loss": 1.8286,
+      "step": 14667
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6380446725490566e-05,
+      "loss": 1.825,
+      "step": 14668
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.637998505054422e-05,
+      "loss": 2.1765,
+      "step": 14669
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6379523352663337e-05,
+      "loss": 1.817,
+      "step": 14670
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6379061631849577e-05,
+      "loss": 1.806,
+      "step": 14671
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.63785998881046e-05,
+      "loss": 1.6328,
+      "step": 14672
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.637813812143006e-05,
+      "loss": 1.8379,
+      "step": 14673
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.637767633182763e-05,
+      "loss": 1.9072,
+      "step": 14674
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6377214519298954e-05,
+      "loss": 1.8266,
+      "step": 14675
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6376752683845705e-05,
+      "loss": 1.849,
+      "step": 14676
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6376290825469535e-05,
+      "loss": 1.9736,
+      "step": 14677
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6375828944172112e-05,
+      "loss": 1.7223,
+      "step": 14678
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6375367039955085e-05,
+      "loss": 2.1333,
+      "step": 14679
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6374905112820125e-05,
+      "loss": 2.0926,
+      "step": 14680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.637444316276889e-05,
+      "loss": 2.0463,
+      "step": 14681
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.637398118980304e-05,
+      "loss": 1.9731,
+      "step": 14682
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6373519193924233e-05,
+      "loss": 1.7633,
+      "step": 14683
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.637305717513413e-05,
+      "loss": 1.6885,
+      "step": 14684
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.6372595133434397e-05,
+      "loss": 1.8834,
+      "step": 14685
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.637213306882669e-05,
+      "loss": 1.8719,
+      "step": 14686
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6371670981312673e-05,
+      "loss": 2.0346,
+      "step": 14687
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6371208870894003e-05,
+      "loss": 1.8988,
+      "step": 14688
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6370746737572345e-05,
+      "loss": 1.6666,
+      "step": 14689
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6370284581349362e-05,
+      "loss": 1.8014,
+      "step": 14690
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.636982240222671e-05,
+      "loss": 1.9252,
+      "step": 14691
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6369360200206053e-05,
+      "loss": 1.6966,
+      "step": 14692
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.636889797528905e-05,
+      "loss": 1.5672,
+      "step": 14693
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6368435727477367e-05,
+      "loss": 2.033,
+      "step": 14694
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6367973456772667e-05,
+      "loss": 2.2015,
+      "step": 14695
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.63675111631766e-05,
+      "loss": 1.6569,
+      "step": 14696
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6367048846690843e-05,
+      "loss": 1.548,
+      "step": 14697
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6366586507317053e-05,
+      "loss": 1.6071,
+      "step": 14698
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6366124145056884e-05,
+      "loss": 2.0378,
+      "step": 14699
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6365661759912005e-05,
+      "loss": 1.9381,
+      "step": 14700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.636519935188408e-05,
+      "loss": 1.858,
+      "step": 14701
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6364736920974768e-05,
+      "loss": 2.0794,
+      "step": 14702
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6364274467185736e-05,
+      "loss": 1.7404,
+      "step": 14703
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6363811990518635e-05,
+      "loss": 1.741,
+      "step": 14704
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.636334949097514e-05,
+      "loss": 2.2637,
+      "step": 14705
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6362886968556907e-05,
+      "loss": 1.6985,
+      "step": 14706
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.63624244232656e-05,
+      "loss": 1.6391,
+      "step": 14707
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.636196185510288e-05,
+      "loss": 1.9716,
+      "step": 14708
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6361499264070415e-05,
+      "loss": 1.7912,
+      "step": 14709
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6361036650169863e-05,
+      "loss": 1.5376,
+      "step": 14710
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6360574013402886e-05,
+      "loss": 1.6619,
+      "step": 14711
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6360111353771157e-05,
+      "loss": 1.9293,
+      "step": 14712
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6359648671276324e-05,
+      "loss": 2.0608,
+      "step": 14713
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6359185965920062e-05,
+      "loss": 1.7447,
+      "step": 14714
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.635872323770403e-05,
+      "loss": 1.7093,
+      "step": 14715
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.635826048662989e-05,
+      "loss": 1.9275,
+      "step": 14716
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6357797712699312e-05,
+      "loss": 1.9369,
+      "step": 14717
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.635733491591395e-05,
+      "loss": 1.9135,
+      "step": 14718
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6356872096275475e-05,
+      "loss": 1.7978,
+      "step": 14719
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6356409253785545e-05,
+      "loss": 2.0597,
+      "step": 14720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.635594638844583e-05,
+      "loss": 1.7438,
+      "step": 14721
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.635548350025799e-05,
+      "loss": 2.1054,
+      "step": 14722
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.635502058922369e-05,
+      "loss": 1.6944,
+      "step": 14723
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6354557655344594e-05,
+      "loss": 2.0258,
+      "step": 14724
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6354094698622363e-05,
+      "loss": 1.6882,
+      "step": 14725
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.635363171905867e-05,
+      "loss": 1.9943,
+      "step": 14726
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.6353168716655166e-05,
+      "loss": 1.6581,
+      "step": 14727
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.635270569141353e-05,
+      "loss": 1.7773,
+      "step": 14728
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6352242643335415e-05,
+      "loss": 1.7262,
+      "step": 14729
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.635177957242249e-05,
+      "loss": 1.8182,
+      "step": 14730
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6351316478676426e-05,
+      "loss": 2.0239,
+      "step": 14731
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6350853362098878e-05,
+      "loss": 1.7799,
+      "step": 14732
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.635039022269151e-05,
+      "loss": 1.8151,
+      "step": 14733
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6349927060455995e-05,
+      "loss": 2.0161,
+      "step": 14734
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.634946387539399e-05,
+      "loss": 2.1656,
+      "step": 14735
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.634900066750717e-05,
+      "loss": 1.9083,
+      "step": 14736
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.634853743679719e-05,
+      "loss": 1.8375,
+      "step": 14737
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.634807418326572e-05,
+      "loss": 2.0124,
+      "step": 14738
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6347610906914422e-05,
+      "loss": 1.6694,
+      "step": 14739
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6347147607744967e-05,
+      "loss": 1.6105,
+      "step": 14740
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.634668428575902e-05,
+      "loss": 1.949,
+      "step": 14741
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.634622094095824e-05,
+      "loss": 1.9557,
+      "step": 14742
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6345757573344297e-05,
+      "loss": 1.7636,
+      "step": 14743
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.634529418291886e-05,
+      "loss": 1.9232,
+      "step": 14744
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6344830769683584e-05,
+      "loss": 2.1856,
+      "step": 14745
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6344367333640144e-05,
+      "loss": 2.0918,
+      "step": 14746
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.634390387479021e-05,
+      "loss": 1.7944,
+      "step": 14747
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.6343440393135438e-05,
+      "loss": 2.6582,
+      "step": 14748
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6342976888677492e-05,
+      "loss": 1.9481,
+      "step": 14749
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6342513361418053e-05,
+      "loss": 1.7242,
+      "step": 14750
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6342049811358777e-05,
+      "loss": 1.7771,
+      "step": 14751
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.6341586238501328e-05,
+      "loss": 1.6389,
+      "step": 14752
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6341122642847377e-05,
+      "loss": 2.3089,
+      "step": 14753
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.634065902439859e-05,
+      "loss": 1.7841,
+      "step": 14754
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6340195383156637e-05,
+      "loss": 1.9175,
+      "step": 14755
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6339731719123176e-05,
+      "loss": 1.8663,
+      "step": 14756
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6339268032299877e-05,
+      "loss": 1.7954,
+      "step": 14757
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6338804322688414e-05,
+      "loss": 1.7988,
+      "step": 14758
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6338340590290446e-05,
+      "loss": 1.6045,
+      "step": 14759
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6337876835107644e-05,
+      "loss": 2.3232,
+      "step": 14760
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6337413057141675e-05,
+      "loss": 1.7221,
+      "step": 14761
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6336949256394203e-05,
+      "loss": 2.194,
+      "step": 14762
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6336485432866895e-05,
+      "loss": 1.8397,
+      "step": 14763
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6336021586561424e-05,
+      "loss": 2.2949,
+      "step": 14764
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.633555771747945e-05,
+      "loss": 1.9056,
+      "step": 14765
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.6335093825622648e-05,
+      "loss": 2.0963,
+      "step": 14766
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.633462991099268e-05,
+      "loss": 1.7002,
+      "step": 14767
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6334165973591218e-05,
+      "loss": 1.9489,
+      "step": 14768
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6333702013419924e-05,
+      "loss": 1.7789,
+      "step": 14769
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6333238030480473e-05,
+      "loss": 2.0768,
+      "step": 14770
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6332774024774524e-05,
+      "loss": 2.1798,
+      "step": 14771
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.633230999630375e-05,
+      "loss": 2.2224,
+      "step": 14772
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6331845945069823e-05,
+      "loss": 1.8971,
+      "step": 14773
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6331381871074408e-05,
+      "loss": 1.7074,
+      "step": 14774
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.633091777431917e-05,
+      "loss": 1.615,
+      "step": 14775
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.633045365480578e-05,
+      "loss": 2.0125,
+      "step": 14776
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.0,
+      "learning_rate": 1.6329989512535907e-05,
+      "loss": 3.3516,
+      "step": 14777
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.632952534751122e-05,
+      "loss": 1.7573,
+      "step": 14778
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6329061159733384e-05,
+      "loss": 1.6775,
+      "step": 14779
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.632859694920407e-05,
+      "loss": 2.0971,
+      "step": 14780
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6328132715924947e-05,
+      "loss": 1.9585,
+      "step": 14781
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6327668459897683e-05,
+      "loss": 2.0566,
+      "step": 14782
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6327204181123948e-05,
+      "loss": 1.8716,
+      "step": 14783
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.632673987960541e-05,
+      "loss": 1.7889,
+      "step": 14784
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6326275555343736e-05,
+      "loss": 1.8501,
+      "step": 14785
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6325811208340603e-05,
+      "loss": 2.1976,
+      "step": 14786
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.632534683859767e-05,
+      "loss": 1.8648,
+      "step": 14787
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.632488244611661e-05,
+      "loss": 1.8595,
+      "step": 14788
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.63244180308991e-05,
+      "loss": 1.6698,
+      "step": 14789
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6323953592946797e-05,
+      "loss": 1.681,
+      "step": 14790
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.632348913226138e-05,
+      "loss": 2.0186,
+      "step": 14791
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6323024648844512e-05,
+      "loss": 2.0268,
+      "step": 14792
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6322560142697868e-05,
+      "loss": 1.8377,
+      "step": 14793
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6322095613823114e-05,
+      "loss": 1.7088,
+      "step": 14794
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6321631062221923e-05,
+      "loss": 1.5448,
+      "step": 14795
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.6321166487895964e-05,
+      "loss": 1.8676,
+      "step": 14796
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6320701890846903e-05,
+      "loss": 1.798,
+      "step": 14797
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6320237271076416e-05,
+      "loss": 2.38,
+      "step": 14798
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.631977262858617e-05,
+      "loss": 1.8682,
+      "step": 14799
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6319307963377836e-05,
+      "loss": 2.1984,
+      "step": 14800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.6318843275453083e-05,
+      "loss": 2.1893,
+      "step": 14801
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6318378564813583e-05,
+      "loss": 1.8236,
+      "step": 14802
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.631791383146101e-05,
+      "loss": 2.2769,
+      "step": 14803
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6317449075397026e-05,
+      "loss": 1.9908,
+      "step": 14804
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6316984296623314e-05,
+      "loss": 1.9,
+      "step": 14805
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.631651949514153e-05,
+      "loss": 1.9228,
+      "step": 14806
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6316054670953353e-05,
+      "loss": 2.0666,
+      "step": 14807
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6315589824060455e-05,
+      "loss": 1.8253,
+      "step": 14808
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6315124954464506e-05,
+      "loss": 1.965,
+      "step": 14809
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6314660062167176e-05,
+      "loss": 1.9513,
+      "step": 14810
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.631419514717013e-05,
+      "loss": 1.8162,
+      "step": 14811
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6313730209475058e-05,
+      "loss": 1.8794,
+      "step": 14812
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.631326524908361e-05,
+      "loss": 1.9594,
+      "step": 14813
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6312800265997467e-05,
+      "loss": 1.9207,
+      "step": 14814
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6312335260218305e-05,
+      "loss": 1.6669,
+      "step": 14815
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6311870231747788e-05,
+      "loss": 1.8199,
+      "step": 14816
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.631140518058759e-05,
+      "loss": 2.2959,
+      "step": 14817
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.631094010673938e-05,
+      "loss": 1.6939,
+      "step": 14818
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.6310475010204837e-05,
+      "loss": 2.2971,
+      "step": 14819
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6310009890985625e-05,
+      "loss": 1.9325,
+      "step": 14820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.630954474908342e-05,
+      "loss": 1.6948,
+      "step": 14821
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.63090795844999e-05,
+      "loss": 1.8841,
+      "step": 14822
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6308614397236728e-05,
+      "loss": 1.7778,
+      "step": 14823
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6308149187295574e-05,
+      "loss": 1.6967,
+      "step": 14824
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6307683954678122e-05,
+      "loss": 1.8484,
+      "step": 14825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.6307218699386032e-05,
+      "loss": 2.0886,
+      "step": 14826
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6306753421420988e-05,
+      "loss": 1.7922,
+      "step": 14827
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6306288120784656e-05,
+      "loss": 1.9611,
+      "step": 14828
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6305822797478707e-05,
+      "loss": 2.0605,
+      "step": 14829
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6305357451504818e-05,
+      "loss": 1.8249,
+      "step": 14830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.630489208286466e-05,
+      "loss": 1.8284,
+      "step": 14831
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6304426691559905e-05,
+      "loss": 2.1145,
+      "step": 14832
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6303961277592228e-05,
+      "loss": 1.9198,
+      "step": 14833
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.63034958409633e-05,
+      "loss": 1.7137,
+      "step": 14834
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6303030381674796e-05,
+      "loss": 1.9967,
+      "step": 14835
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.630256489972839e-05,
+      "loss": 1.79,
+      "step": 14836
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.630209939512575e-05,
+      "loss": 1.5491,
+      "step": 14837
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6301633867868555e-05,
+      "loss": 2.0461,
+      "step": 14838
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6301168317958477e-05,
+      "loss": 1.9574,
+      "step": 14839
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.630070274539719e-05,
+      "loss": 1.9885,
+      "step": 14840
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6300237150186363e-05,
+      "loss": 1.7713,
+      "step": 14841
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6299771532327678e-05,
+      "loss": 2.148,
+      "step": 14842
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.62993058918228e-05,
+      "loss": 1.8679,
+      "step": 14843
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.629884022867341e-05,
+      "loss": 2.011,
+      "step": 14844
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.629837454288118e-05,
+      "loss": 1.9412,
+      "step": 14845
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.629790883444778e-05,
+      "loss": 1.6285,
+      "step": 14846
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6297443103374888e-05,
+      "loss": 1.5954,
+      "step": 14847
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.6296977349664176e-05,
+      "loss": 1.9224,
+      "step": 14848
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.6296511573317322e-05,
+      "loss": 1.6315,
+      "step": 14849
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6296045774335997e-05,
+      "loss": 1.9862,
+      "step": 14850
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6295579952721876e-05,
+      "loss": 1.7957,
+      "step": 14851
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6295114108476634e-05,
+      "loss": 1.8966,
+      "step": 14852
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6294648241601948e-05,
+      "loss": 1.9096,
+      "step": 14853
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.629418235209949e-05,
+      "loss": 1.9214,
+      "step": 14854
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6293716439970932e-05,
+      "loss": 1.7988,
+      "step": 14855
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6293250505217953e-05,
+      "loss": 1.7847,
+      "step": 14856
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.6292784547842225e-05,
+      "loss": 2.2708,
+      "step": 14857
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.629231856784543e-05,
+      "loss": 1.7143,
+      "step": 14858
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6291852565229234e-05,
+      "loss": 1.8426,
+      "step": 14859
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6291386539995315e-05,
+      "loss": 2.1409,
+      "step": 14860
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.629092049214535e-05,
+      "loss": 1.9727,
+      "step": 14861
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6290454421681014e-05,
+      "loss": 1.8048,
+      "step": 14862
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6289988328603984e-05,
+      "loss": 1.9008,
+      "step": 14863
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.628952221291593e-05,
+      "loss": 1.6265,
+      "step": 14864
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6289056074618533e-05,
+      "loss": 1.5679,
+      "step": 14865
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.628858991371347e-05,
+      "loss": 1.7461,
+      "step": 14866
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6288123730202408e-05,
+      "loss": 1.8191,
+      "step": 14867
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.628765752408703e-05,
+      "loss": 1.815,
+      "step": 14868
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6287191295369012e-05,
+      "loss": 1.9811,
+      "step": 14869
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6286725044050026e-05,
+      "loss": 2.0609,
+      "step": 14870
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.628625877013175e-05,
+      "loss": 1.7232,
+      "step": 14871
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.628579247361586e-05,
+      "loss": 2.1827,
+      "step": 14872
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6285326154504037e-05,
+      "loss": 2.0872,
+      "step": 14873
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6284859812797947e-05,
+      "loss": 1.6066,
+      "step": 14874
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6284393448499277e-05,
+      "loss": 1.7731,
+      "step": 14875
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6283927061609696e-05,
+      "loss": 1.6632,
+      "step": 14876
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6283460652130884e-05,
+      "loss": 1.8272,
+      "step": 14877
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.628299422006452e-05,
+      "loss": 1.962,
+      "step": 14878
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.6282527765412272e-05,
+      "loss": 2.1211,
+      "step": 14879
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6282061288175826e-05,
+      "loss": 1.595,
+      "step": 14880
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0,
+      "learning_rate": 1.6281594788356854e-05,
+      "loss": 1.7106,
+      "step": 14881
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6281128265957038e-05,
+      "loss": 2.045,
+      "step": 14882
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6280661720978048e-05,
+      "loss": 1.9146,
+      "step": 14883
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6280195153421563e-05,
+      "loss": 1.8061,
+      "step": 14884
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6279728563289262e-05,
+      "loss": 1.7928,
+      "step": 14885
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6279261950582825e-05,
+      "loss": 1.8829,
+      "step": 14886
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6278795315303922e-05,
+      "loss": 1.7779,
+      "step": 14887
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.627832865745424e-05,
+      "loss": 2.0843,
+      "step": 14888
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.627786197703545e-05,
+      "loss": 1.813,
+      "step": 14889
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6277395274049227e-05,
+      "loss": 1.5374,
+      "step": 14890
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6276928548497256e-05,
+      "loss": 1.9006,
+      "step": 14891
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.627646180038121e-05,
+      "loss": 1.8708,
+      "step": 14892
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.627599502970277e-05,
+      "loss": 1.7961,
+      "step": 14893
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.627552823646361e-05,
+      "loss": 1.7116,
+      "step": 14894
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.627506142066541e-05,
+      "loss": 1.9923,
+      "step": 14895
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6274594582309847e-05,
+      "loss": 1.8304,
+      "step": 14896
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6274127721398606e-05,
+      "loss": 1.804,
+      "step": 14897
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6273660837933354e-05,
+      "loss": 1.8608,
+      "step": 14898
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.627319393191578e-05,
+      "loss": 2.0419,
+      "step": 14899
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6272727003347554e-05,
+      "loss": 1.9722,
+      "step": 14900
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6272260052230356e-05,
+      "loss": 2.0974,
+      "step": 14901
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.627179307856587e-05,
+      "loss": 1.963,
+      "step": 14902
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6271326082355766e-05,
+      "loss": 1.8375,
+      "step": 14903
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6270859063601734e-05,
+      "loss": 2.0384,
+      "step": 14904
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6270392022305444e-05,
+      "loss": 1.8137,
+      "step": 14905
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.626992495846858e-05,
+      "loss": 1.6019,
+      "step": 14906
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6269457872092816e-05,
+      "loss": 1.6695,
+      "step": 14907
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.6268990763179832e-05,
+      "loss": 1.9252,
+      "step": 14908
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6268523631731312e-05,
+      "loss": 2.1546,
+      "step": 14909
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.626805647774893e-05,
+      "loss": 1.6751,
+      "step": 14910
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.626758930123437e-05,
+      "loss": 1.651,
+      "step": 14911
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6267122102189306e-05,
+      "loss": 1.8459,
+      "step": 14912
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.626665488061542e-05,
+      "loss": 2.0894,
+      "step": 14913
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6266187636514392e-05,
+      "loss": 2.7236,
+      "step": 14914
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6265720369887904e-05,
+      "loss": 2.0094,
+      "step": 14915
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6265253080737632e-05,
+      "loss": 1.6545,
+      "step": 14916
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6264785769065255e-05,
+      "loss": 2.1074,
+      "step": 14917
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6264318434872453e-05,
+      "loss": 2.028,
+      "step": 14918
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6263851078160912e-05,
+      "loss": 2.0125,
+      "step": 14919
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.6263383698932307e-05,
+      "loss": 1.7671,
+      "step": 14920
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6262916297188316e-05,
+      "loss": 1.8051,
+      "step": 14921
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6262448872930624e-05,
+      "loss": 1.6266,
+      "step": 14922
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.6261981426160913e-05,
+      "loss": 2.1536,
+      "step": 14923
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6261513956880858e-05,
+      "loss": 1.8492,
+      "step": 14924
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6261046465092135e-05,
+      "loss": 2.0287,
+      "step": 14925
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6260578950796436e-05,
+      "loss": 1.6519,
+      "step": 14926
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6260111413995434e-05,
+      "loss": 1.8557,
+      "step": 14927
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6259643854690813e-05,
+      "loss": 1.9006,
+      "step": 14928
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6259176272884257e-05,
+      "loss": 1.8096,
+      "step": 14929
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.6258708668577436e-05,
+      "loss": 1.8582,
+      "step": 14930
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.625824104177204e-05,
+      "loss": 2.1024,
+      "step": 14931
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.6257773392469746e-05,
+      "loss": 2.4508,
+      "step": 14932
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6257305720672238e-05,
+      "loss": 1.7364,
+      "step": 14933
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6256838026381197e-05,
+      "loss": 1.9798,
+      "step": 14934
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.62563703095983e-05,
+      "loss": 1.7903,
+      "step": 14935
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6255902570325237e-05,
+      "loss": 1.9829,
+      "step": 14936
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.625543480856368e-05,
+      "loss": 1.674,
+      "step": 14937
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.625496702431531e-05,
+      "loss": 2.3372,
+      "step": 14938
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.625449921758182e-05,
+      "loss": 1.8997,
+      "step": 14939
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.625403138836488e-05,
+      "loss": 1.9428,
+      "step": 14940
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6253563536666176e-05,
+      "loss": 1.7059,
+      "step": 14941
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6253095662487394e-05,
+      "loss": 2.2232,
+      "step": 14942
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6252627765830207e-05,
+      "loss": 1.9389,
+      "step": 14943
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6252159846696304e-05,
+      "loss": 1.8663,
+      "step": 14944
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6251691905087367e-05,
+      "loss": 2.0624,
+      "step": 14945
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6251223941005074e-05,
+      "loss": 1.8728,
+      "step": 14946
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6250755954451108e-05,
+      "loss": 1.7668,
+      "step": 14947
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6250287945427158e-05,
+      "loss": 1.9313,
+      "step": 14948
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6249819913934896e-05,
+      "loss": 1.8874,
+      "step": 14949
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.624935185997601e-05,
+      "loss": 1.8795,
+      "step": 14950
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.624888378355218e-05,
+      "loss": 1.935,
+      "step": 14951
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6248415684665096e-05,
+      "loss": 1.6966,
+      "step": 14952
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6247947563316432e-05,
+      "loss": 2.1964,
+      "step": 14953
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6247479419507876e-05,
+      "loss": 1.7848,
+      "step": 14954
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.624701125324111e-05,
+      "loss": 2.0006,
+      "step": 14955
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.624654306451781e-05,
+      "loss": 1.7833,
+      "step": 14956
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.624607485333967e-05,
+      "loss": 1.6577,
+      "step": 14957
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.624560661970836e-05,
+      "loss": 2.0562,
+      "step": 14958
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6245138363625585e-05,
+      "loss": 1.8213,
+      "step": 14959
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6244670085093006e-05,
+      "loss": 1.7005,
+      "step": 14960
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6244201784112313e-05,
+      "loss": 1.7575,
+      "step": 14961
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6243733460685196e-05,
+      "loss": 1.8329,
+      "step": 14962
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6243265114813332e-05,
+      "loss": 1.7479,
+      "step": 14963
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.624279674649841e-05,
+      "loss": 1.7317,
+      "step": 14964
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.6242328355742107e-05,
+      "loss": 1.917,
+      "step": 14965
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.624185994254611e-05,
+      "loss": 1.8567,
+      "step": 14966
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6241391506912097e-05,
+      "loss": 2.2729,
+      "step": 14967
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6240923048841765e-05,
+      "loss": 1.7523,
+      "step": 14968
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.624045456833679e-05,
+      "loss": 2.0342,
+      "step": 14969
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6239986065398854e-05,
+      "loss": 1.7794,
+      "step": 14970
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6239517540029646e-05,
+      "loss": 1.9088,
+      "step": 14971
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6239048992230847e-05,
+      "loss": 1.6075,
+      "step": 14972
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6238580422004143e-05,
+      "loss": 1.943,
+      "step": 14973
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6238111829351215e-05,
+      "loss": 1.9828,
+      "step": 14974
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.623764321427375e-05,
+      "loss": 2.2542,
+      "step": 14975
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6237174576773438e-05,
+      "loss": 1.6924,
+      "step": 14976
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6236705916851956e-05,
+      "loss": 1.7195,
+      "step": 14977
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.623623723451099e-05,
+      "loss": 1.9043,
+      "step": 14978
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6235768529752224e-05,
+      "loss": 1.6711,
+      "step": 14979
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6235299802577352e-05,
+      "loss": 1.5933,
+      "step": 14980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6234831052988044e-05,
+      "loss": 1.6169,
+      "step": 14981
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6234362280985997e-05,
+      "loss": 1.8755,
+      "step": 14982
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6233893486572894e-05,
+      "loss": 1.825,
+      "step": 14983
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6233424669750413e-05,
+      "loss": 1.9051,
+      "step": 14984
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.623295583052025e-05,
+      "loss": 1.7217,
+      "step": 14985
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.623248696888408e-05,
+      "loss": 1.9896,
+      "step": 14986
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6232018084843596e-05,
+      "loss": 1.7621,
+      "step": 14987
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6231549178400476e-05,
+      "loss": 1.7943,
+      "step": 14988
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6231080249556414e-05,
+      "loss": 1.6777,
+      "step": 14989
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6230611298313095e-05,
+      "loss": 1.8288,
+      "step": 14990
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6230142324672198e-05,
+      "loss": 2.1691,
+      "step": 14991
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6229673328635415e-05,
+      "loss": 1.7719,
+      "step": 14992
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.622920431020443e-05,
+      "loss": 2.2586,
+      "step": 14993
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.6228735269380928e-05,
+      "loss": 1.719,
+      "step": 14994
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.622826620616659e-05,
+      "loss": 1.8207,
+      "step": 14995
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6227797120563115e-05,
+      "loss": 2.1458,
+      "step": 14996
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.622732801257218e-05,
+      "loss": 1.8032,
+      "step": 14997
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6226858882195474e-05,
+      "loss": 2.0368,
+      "step": 14998
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6226389729434685e-05,
+      "loss": 1.8976,
+      "step": 14999
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6225920554291494e-05,
+      "loss": 1.7165,
+      "step": 15000
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.622545135676759e-05,
+      "loss": 1.4429,
+      "step": 15001
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6224982136864663e-05,
+      "loss": 1.8855,
+      "step": 15002
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6224512894584397e-05,
+      "loss": 2.2319,
+      "step": 15003
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.622404362992848e-05,
+      "loss": 1.9015,
+      "step": 15004
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.62235743428986e-05,
+      "loss": 1.7669,
+      "step": 15005
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6223105033496437e-05,
+      "loss": 2.1527,
+      "step": 15006
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6222635701723685e-05,
+      "loss": 1.985,
+      "step": 15007
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.622216634758203e-05,
+      "loss": 2.0393,
+      "step": 15008
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6221696971073157e-05,
+      "loss": 1.7006,
+      "step": 15009
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.622122757219876e-05,
+      "loss": 1.9843,
+      "step": 15010
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6220758150960513e-05,
+      "loss": 1.9403,
+      "step": 15011
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.6220288707360116e-05,
+      "loss": 1.7495,
+      "step": 15012
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6219819241399256e-05,
+      "loss": 1.7332,
+      "step": 15013
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6219349753079612e-05,
+      "loss": 2.0302,
+      "step": 15014
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6218880242402876e-05,
+      "loss": 1.8464,
+      "step": 15015
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6218410709370735e-05,
+      "loss": 1.9436,
+      "step": 15016
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6217941153984882e-05,
+      "loss": 1.7884,
+      "step": 15017
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6217471576247003e-05,
+      "loss": 1.9156,
+      "step": 15018
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.6217001976158777e-05,
+      "loss": 2.0632,
+      "step": 15019
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6216532353721905e-05,
+      "loss": 1.8328,
+      "step": 15020
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6216062708938067e-05,
+      "loss": 1.9727,
+      "step": 15021
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6215593041808952e-05,
+      "loss": 1.9471,
+      "step": 15022
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6215123352336256e-05,
+      "loss": 1.7815,
+      "step": 15023
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6214653640521657e-05,
+      "loss": 1.7027,
+      "step": 15024
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6214183906366848e-05,
+      "loss": 2.0,
+      "step": 15025
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.6213714149873517e-05,
+      "loss": 2.0639,
+      "step": 15026
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.621324437104335e-05,
+      "loss": 1.9079,
+      "step": 15027
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6212774569878045e-05,
+      "loss": 2.2132,
+      "step": 15028
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6212304746379282e-05,
+      "loss": 1.8524,
+      "step": 15029
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6211834900548752e-05,
+      "loss": 1.8614,
+      "step": 15030
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6211365032388147e-05,
+      "loss": 1.8874,
+      "step": 15031
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.621089514189915e-05,
+      "loss": 1.8212,
+      "step": 15032
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6210425229083457e-05,
+      "loss": 1.9103,
+      "step": 15033
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6209955293942753e-05,
+      "loss": 2.0878,
+      "step": 15034
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.620948533647873e-05,
+      "loss": 1.7404,
+      "step": 15035
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6209015356693073e-05,
+      "loss": 1.7618,
+      "step": 15036
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6208545354587476e-05,
+      "loss": 2.0084,
+      "step": 15037
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6208075330163624e-05,
+      "loss": 2.0288,
+      "step": 15038
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6207605283423212e-05,
+      "loss": 1.9017,
+      "step": 15039
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6207135214367927e-05,
+      "loss": 2.0783,
+      "step": 15040
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.620666512299946e-05,
+      "loss": 1.9096,
+      "step": 15041
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.6206195009319497e-05,
+      "loss": 2.1783,
+      "step": 15042
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.6205724873329728e-05,
+      "loss": 1.7387,
+      "step": 15043
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.6205254715031854e-05,
+      "loss": 1.6511,
+      "step": 15044
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6204784534427548e-05,
+      "loss": 1.8304,
+      "step": 15045
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6204314331518516e-05,
+      "loss": 1.7443,
+      "step": 15046
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.620384410630644e-05,
+      "loss": 1.8825,
+      "step": 15047
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6203373858793008e-05,
+      "loss": 1.8985,
+      "step": 15048
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.6202903588979917e-05,
+      "loss": 1.9165,
+      "step": 15049
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6202433296868852e-05,
+      "loss": 1.5149,
+      "step": 15050
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.620196298246151e-05,
+      "loss": 1.8871,
+      "step": 15051
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6201492645759574e-05,
+      "loss": 1.5314,
+      "step": 15052
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.620102228676474e-05,
+      "loss": 2.1822,
+      "step": 15053
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.62005519054787e-05,
+      "loss": 1.816,
+      "step": 15054
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.6200081501903134e-05,
+      "loss": 1.9942,
+      "step": 15055
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.619961107603975e-05,
+      "loss": 1.973,
+      "step": 15056
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6199140627890228e-05,
+      "loss": 1.899,
+      "step": 15057
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.619867015745626e-05,
+      "loss": 1.7857,
+      "step": 15058
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.6198199664739542e-05,
+      "loss": 1.2133,
+      "step": 15059
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.619772914974176e-05,
+      "loss": 1.7146,
+      "step": 15060
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6197258612464604e-05,
+      "loss": 2.0425,
+      "step": 15061
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6196788052909772e-05,
+      "loss": 1.7801,
+      "step": 15062
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6196317471078952e-05,
+      "loss": 1.632,
+      "step": 15063
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6195846866973835e-05,
+      "loss": 1.9216,
+      "step": 15064
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6195376240596116e-05,
+      "loss": 1.8237,
+      "step": 15065
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6194905591947484e-05,
+      "loss": 1.9948,
+      "step": 15066
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6194434921029632e-05,
+      "loss": 1.7482,
+      "step": 15067
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.619396422784425e-05,
+      "loss": 2.1469,
+      "step": 15068
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6193493512393033e-05,
+      "loss": 2.0179,
+      "step": 15069
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.619302277467767e-05,
+      "loss": 2.3622,
+      "step": 15070
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6192552014699857e-05,
+      "loss": 1.9197,
+      "step": 15071
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6192081232461282e-05,
+      "loss": 1.8164,
+      "step": 15072
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.619161042796364e-05,
+      "loss": 1.7255,
+      "step": 15073
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6191139601208622e-05,
+      "loss": 1.9923,
+      "step": 15074
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.619066875219792e-05,
+      "loss": 2.0308,
+      "step": 15075
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.619019788093323e-05,
+      "loss": 1.7398,
+      "step": 15076
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6189726987416244e-05,
+      "loss": 2.0067,
+      "step": 15077
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.618925607164865e-05,
+      "loss": 1.9328,
+      "step": 15078
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6188785133632145e-05,
+      "loss": 1.8035,
+      "step": 15079
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6188314173368424e-05,
+      "loss": 1.5182,
+      "step": 15080
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6187843190859175e-05,
+      "loss": 2.0424,
+      "step": 15081
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.618737218610609e-05,
+      "loss": 1.7025,
+      "step": 15082
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6186901159110872e-05,
+      "loss": 2.1175,
+      "step": 15083
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6186430109875203e-05,
+      "loss": 2.0136,
+      "step": 15084
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.618595903840078e-05,
+      "loss": 1.8593,
+      "step": 15085
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6185487944689297e-05,
+      "loss": 1.8313,
+      "step": 15086
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.618501682874245e-05,
+      "loss": 1.805,
+      "step": 15087
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.618454569056193e-05,
+      "loss": 2.1791,
+      "step": 15088
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.618407453014943e-05,
+      "loss": 1.8769,
+      "step": 15089
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6183603347506644e-05,
+      "loss": 1.9923,
+      "step": 15090
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6183132142635268e-05,
+      "loss": 1.9494,
+      "step": 15091
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6182660915536995e-05,
+      "loss": 2.0116,
+      "step": 15092
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6182189666213516e-05,
+      "loss": 2.2316,
+      "step": 15093
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.618171839466653e-05,
+      "loss": 1.9466,
+      "step": 15094
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.6181247100897723e-05,
+      "loss": 1.8708,
+      "step": 15095
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6180775784908797e-05,
+      "loss": 1.7596,
+      "step": 15096
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.618030444670145e-05,
+      "loss": 2.4515,
+      "step": 15097
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.617983308627736e-05,
+      "loss": 1.7298,
+      "step": 15098
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.6179361703638238e-05,
+      "loss": 1.8721,
+      "step": 15099
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6178890298785767e-05,
+      "loss": 2.0764,
+      "step": 15100
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.617841887172165e-05,
+      "loss": 1.9084,
+      "step": 15101
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.617794742244758e-05,
+      "loss": 1.8062,
+      "step": 15102
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6177475950965246e-05,
+      "loss": 1.9157,
+      "step": 15103
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.617700445727635e-05,
+      "loss": 1.9229,
+      "step": 15104
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6176532941382583e-05,
+      "loss": 1.8985,
+      "step": 15105
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.617606140328564e-05,
+      "loss": 1.9534,
+      "step": 15106
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.617558984298722e-05,
+      "loss": 1.9662,
+      "step": 15107
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.617511826048901e-05,
+      "loss": 1.9328,
+      "step": 15108
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6174646655792713e-05,
+      "loss": 1.6876,
+      "step": 15109
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6174175028900022e-05,
+      "loss": 1.8596,
+      "step": 15110
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6173703379812632e-05,
+      "loss": 1.873,
+      "step": 15111
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6173231708532236e-05,
+      "loss": 1.9369,
+      "step": 15112
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6172760015060533e-05,
+      "loss": 1.8698,
+      "step": 15113
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6172288299399223e-05,
+      "loss": 1.8889,
+      "step": 15114
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.617181656154999e-05,
+      "loss": 2.3191,
+      "step": 15115
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6171344801514534e-05,
+      "loss": 1.8025,
+      "step": 15116
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6170873019294558e-05,
+      "loss": 1.6431,
+      "step": 15117
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.617040121489175e-05,
+      "loss": 1.7214,
+      "step": 15118
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.616992938830781e-05,
+      "loss": 1.728,
+      "step": 15119
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6169457539544433e-05,
+      "loss": 1.7057,
+      "step": 15120
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6168985668603315e-05,
+      "loss": 1.9791,
+      "step": 15121
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.6168513775486153e-05,
+      "loss": 1.9376,
+      "step": 15122
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6168041860194644e-05,
+      "loss": 2.1324,
+      "step": 15123
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.616756992273048e-05,
+      "loss": 1.9308,
+      "step": 15124
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6167097963095365e-05,
+      "loss": 2.0083,
+      "step": 15125
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6166625981290987e-05,
+      "loss": 1.967,
+      "step": 15126
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6166153977319048e-05,
+      "loss": 1.97,
+      "step": 15127
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6165681951181243e-05,
+      "loss": 1.9169,
+      "step": 15128
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6165209902879273e-05,
+      "loss": 1.9737,
+      "step": 15129
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6164737832414828e-05,
+      "loss": 1.9517,
+      "step": 15130
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6164265739789612e-05,
+      "loss": 1.7583,
+      "step": 15131
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6163793625005314e-05,
+      "loss": 1.7262,
+      "step": 15132
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6163321488063636e-05,
+      "loss": 1.6302,
+      "step": 15133
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.616284932896628e-05,
+      "loss": 1.7613,
+      "step": 15134
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6162377147714933e-05,
+      "loss": 2.0679,
+      "step": 15135
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.61619049443113e-05,
+      "loss": 1.8859,
+      "step": 15136
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6161432718757077e-05,
+      "loss": 2.1378,
+      "step": 15137
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.616096047105396e-05,
+      "loss": 1.8522,
+      "step": 15138
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6160488201203643e-05,
+      "loss": 2.0801,
+      "step": 15139
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6160015909207827e-05,
+      "loss": 1.9775,
+      "step": 15140
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.615954359506822e-05,
+      "loss": 1.988,
+      "step": 15141
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6159071258786502e-05,
+      "loss": 1.9371,
+      "step": 15142
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6158598900364384e-05,
+      "loss": 1.9149,
+      "step": 15143
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6158126519803555e-05,
+      "loss": 2.0842,
+      "step": 15144
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6157654117105724e-05,
+      "loss": 1.5362,
+      "step": 15145
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6157181692272577e-05,
+      "loss": 1.6965,
+      "step": 15146
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.615670924530582e-05,
+      "loss": 1.801,
+      "step": 15147
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.615623677620715e-05,
+      "loss": 2.0771,
+      "step": 15148
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6155764284978262e-05,
+      "loss": 1.5866,
+      "step": 15149
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.615529177162086e-05,
+      "loss": 1.9313,
+      "step": 15150
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.615481923613664e-05,
+      "loss": 1.8397,
+      "step": 15151
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.6154346678527297e-05,
+      "loss": 2.2204,
+      "step": 15152
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6153874098794535e-05,
+      "loss": 1.7821,
+      "step": 15153
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6153401496940054e-05,
+      "loss": 2.0214,
+      "step": 15154
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6152928872965544e-05,
+      "loss": 1.4224,
+      "step": 15155
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6152456226872714e-05,
+      "loss": 1.9004,
+      "step": 15156
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6151983558663258e-05,
+      "loss": 1.5126,
+      "step": 15157
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6151510868338875e-05,
+      "loss": 1.9699,
+      "step": 15158
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6151038155901265e-05,
+      "loss": 1.7715,
+      "step": 15159
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.615056542135213e-05,
+      "loss": 2.0709,
+      "step": 15160
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6150092664693165e-05,
+      "loss": 1.7471,
+      "step": 15161
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.614961988592607e-05,
+      "loss": 1.6021,
+      "step": 15162
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.614914708505255e-05,
+      "loss": 1.7293,
+      "step": 15163
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6148674262074297e-05,
+      "loss": 1.9919,
+      "step": 15164
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6148201416993017e-05,
+      "loss": 1.8524,
+      "step": 15165
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6147728549810405e-05,
+      "loss": 2.1381,
+      "step": 15166
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.614725566052816e-05,
+      "loss": 1.9589,
+      "step": 15167
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.614678274914799e-05,
+      "loss": 2.0699,
+      "step": 15168
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.614630981567159e-05,
+      "loss": 1.9726,
+      "step": 15169
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6145836860100655e-05,
+      "loss": 1.6377,
+      "step": 15170
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.614536388243689e-05,
+      "loss": 2.0397,
+      "step": 15171
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6144890882681997e-05,
+      "loss": 1.9127,
+      "step": 15172
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6144417860837677e-05,
+      "loss": 1.4261,
+      "step": 15173
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6143944816905625e-05,
+      "loss": 1.7159,
+      "step": 15174
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6143471750887548e-05,
+      "loss": 2.102,
+      "step": 15175
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6142998662785137e-05,
+      "loss": 1.7359,
+      "step": 15176
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6142525552600102e-05,
+      "loss": 1.6603,
+      "step": 15177
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.614205242033414e-05,
+      "loss": 2.0865,
+      "step": 15178
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6141579265988953e-05,
+      "loss": 1.9198,
+      "step": 15179
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.614110608956624e-05,
+      "loss": 1.948,
+      "step": 15180
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6140632891067703e-05,
+      "loss": 2.0888,
+      "step": 15181
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.614015967049504e-05,
+      "loss": 1.8694,
+      "step": 15182
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.6139686427849958e-05,
+      "loss": 2.0088,
+      "step": 15183
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6139213163134156e-05,
+      "loss": 2.1458,
+      "step": 15184
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6138739876349335e-05,
+      "loss": 1.9246,
+      "step": 15185
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6138266567497195e-05,
+      "loss": 1.8976,
+      "step": 15186
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6137793236579437e-05,
+      "loss": 1.7279,
+      "step": 15187
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6137319883597765e-05,
+      "loss": 1.9399,
+      "step": 15188
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.613684650855388e-05,
+      "loss": 1.9963,
+      "step": 15189
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.6136373111449478e-05,
+      "loss": 1.9939,
+      "step": 15190
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6135899692286275e-05,
+      "loss": 1.7224,
+      "step": 15191
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.613542625106596e-05,
+      "loss": 1.6034,
+      "step": 15192
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6134952787790236e-05,
+      "loss": 2.0789,
+      "step": 15193
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.6134479302460807e-05,
+      "loss": 1.8692,
+      "step": 15194
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.613400579507938e-05,
+      "loss": 1.916,
+      "step": 15195
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.613353226564765e-05,
+      "loss": 2.0422,
+      "step": 15196
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6133058714167323e-05,
+      "loss": 1.7778,
+      "step": 15197
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.61325851406401e-05,
+      "loss": 1.8958,
+      "step": 15198
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6132111545067685e-05,
+      "loss": 2.3442,
+      "step": 15199
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6131637927451778e-05,
+      "loss": 1.9563,
+      "step": 15200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6131164287794084e-05,
+      "loss": 1.8625,
+      "step": 15201
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.6130690626096303e-05,
+      "loss": 2.2645,
+      "step": 15202
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6130216942360136e-05,
+      "loss": 1.837,
+      "step": 15203
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.6129743236587293e-05,
+      "loss": 1.9044,
+      "step": 15204
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6129269508779474e-05,
+      "loss": 1.5581,
+      "step": 15205
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6128795758938377e-05,
+      "loss": 1.9277,
+      "step": 15206
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6128321987065708e-05,
+      "loss": 1.6049,
+      "step": 15207
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6127848193163176e-05,
+      "loss": 1.6881,
+      "step": 15208
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6127374377232475e-05,
+      "loss": 1.7987,
+      "step": 15209
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6126900539275313e-05,
+      "loss": 1.8222,
+      "step": 15210
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6126426679293393e-05,
+      "loss": 2.123,
+      "step": 15211
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6125952797288416e-05,
+      "loss": 1.957,
+      "step": 15212
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.612547889326209e-05,
+      "loss": 1.9231,
+      "step": 15213
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6125004967216114e-05,
+      "loss": 1.6878,
+      "step": 15214
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6124531019152194e-05,
+      "loss": 1.9083,
+      "step": 15215
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6124057049072037e-05,
+      "loss": 2.1532,
+      "step": 15216
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6123583056977338e-05,
+      "loss": 1.7152,
+      "step": 15217
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6123109042869807e-05,
+      "loss": 1.8397,
+      "step": 15218
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6122635006751147e-05,
+      "loss": 2.1272,
+      "step": 15219
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6122160948623067e-05,
+      "loss": 1.8388,
+      "step": 15220
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.6121686868487262e-05,
+      "loss": 1.9505,
+      "step": 15221
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6121212766345443e-05,
+      "loss": 2.0409,
+      "step": 15222
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.612073864219931e-05,
+      "loss": 1.6978,
+      "step": 15223
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6120264496050565e-05,
+      "loss": 1.6712,
+      "step": 15224
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6119790327900925e-05,
+      "loss": 1.5906,
+      "step": 15225
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6119316137752078e-05,
+      "loss": 1.8695,
+      "step": 15226
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.611884192560574e-05,
+      "loss": 2.1276,
+      "step": 15227
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6118367691463615e-05,
+      "loss": 1.5824,
+      "step": 15228
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.61178934353274e-05,
+      "loss": 1.7313,
+      "step": 15229
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.611741915719881e-05,
+      "loss": 1.655,
+      "step": 15230
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.611694485707954e-05,
+      "loss": 2.1299,
+      "step": 15231
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6116470534971304e-05,
+      "loss": 1.8292,
+      "step": 15232
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.61159961908758e-05,
+      "loss": 1.8397,
+      "step": 15233
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6115521824794737e-05,
+      "loss": 1.7807,
+      "step": 15234
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.611504743672982e-05,
+      "loss": 1.9857,
+      "step": 15235
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6114573026682757e-05,
+      "loss": 2.0047,
+      "step": 15236
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6114098594655242e-05,
+      "loss": 1.9239,
+      "step": 15237
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6113624140648994e-05,
+      "loss": 1.7031,
+      "step": 15238
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6113149664665714e-05,
+      "loss": 1.8837,
+      "step": 15239
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6112675166707103e-05,
+      "loss": 1.6484,
+      "step": 15240
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6112200646774875e-05,
+      "loss": 1.6107,
+      "step": 15241
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6111726104870724e-05,
+      "loss": 1.7756,
+      "step": 15242
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6111251540996368e-05,
+      "loss": 1.7932,
+      "step": 15243
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6110776955153507e-05,
+      "loss": 1.8418,
+      "step": 15244
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.611030234734385e-05,
+      "loss": 1.8174,
+      "step": 15245
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.61098277175691e-05,
+      "loss": 1.8458,
+      "step": 15246
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.610935306583096e-05,
+      "loss": 2.2127,
+      "step": 15247
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6108878392131146e-05,
+      "loss": 1.5508,
+      "step": 15248
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6108403696471358e-05,
+      "loss": 1.7985,
+      "step": 15249
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6107928978853302e-05,
+      "loss": 1.7754,
+      "step": 15250
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6107454239278685e-05,
+      "loss": 1.9455,
+      "step": 15251
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.6106979477749218e-05,
+      "loss": 2.2839,
+      "step": 15252
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.61065046942666e-05,
+      "loss": 1.6247,
+      "step": 15253
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6106029888832543e-05,
+      "loss": 1.8798,
+      "step": 15254
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.610555506144875e-05,
+      "loss": 1.9787,
+      "step": 15255
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6105080212116936e-05,
+      "loss": 2.0129,
+      "step": 15256
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.61046053408388e-05,
+      "loss": 1.767,
+      "step": 15257
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.610413044761605e-05,
+      "loss": 1.763,
+      "step": 15258
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6103655532450396e-05,
+      "loss": 1.6742,
+      "step": 15259
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6103180595343545e-05,
+      "loss": 2.0063,
+      "step": 15260
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.61027056362972e-05,
+      "loss": 1.9357,
+      "step": 15261
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6102230655313076e-05,
+      "loss": 1.9176,
+      "step": 15262
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.610175565239287e-05,
+      "loss": 1.8556,
+      "step": 15263
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.61012806275383e-05,
+      "loss": 2.1156,
+      "step": 15264
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.6100805580751068e-05,
+      "loss": 2.0391,
+      "step": 15265
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6100330512032885e-05,
+      "loss": 1.8011,
+      "step": 15266
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6099855421385452e-05,
+      "loss": 2.0197,
+      "step": 15267
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.609938030881048e-05,
+      "loss": 1.6855,
+      "step": 15268
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6098905174309683e-05,
+      "loss": 1.8863,
+      "step": 15269
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.6098430017884766e-05,
+      "loss": 2.4752,
+      "step": 15270
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6097954839537432e-05,
+      "loss": 2.1707,
+      "step": 15271
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6097479639269395e-05,
+      "loss": 1.9633,
+      "step": 15272
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6097004417082357e-05,
+      "loss": 1.8816,
+      "step": 15273
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.609652917297803e-05,
+      "loss": 2.1562,
+      "step": 15274
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6096053906958124e-05,
+      "loss": 1.6923,
+      "step": 15275
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6095578619024345e-05,
+      "loss": 2.3214,
+      "step": 15276
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6095103309178402e-05,
+      "loss": 1.8204,
+      "step": 15277
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.6094627977422005e-05,
+      "loss": 1.8506,
+      "step": 15278
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6094152623756866e-05,
+      "loss": 1.6992,
+      "step": 15279
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.6093677248184684e-05,
+      "loss": 1.6561,
+      "step": 15280
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.6093201850707173e-05,
+      "loss": 1.9542,
+      "step": 15281
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6092726431326043e-05,
+      "loss": 2.2757,
+      "step": 15282
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6092250990043003e-05,
+      "loss": 1.8755,
+      "step": 15283
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6091775526859764e-05,
+      "loss": 1.95,
+      "step": 15284
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.609130004177803e-05,
+      "loss": 1.7759,
+      "step": 15285
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6090824534799514e-05,
+      "loss": 2.0436,
+      "step": 15286
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6090349005925925e-05,
+      "loss": 1.8243,
+      "step": 15287
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6089873455158968e-05,
+      "loss": 1.994,
+      "step": 15288
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.608939788250036e-05,
+      "loss": 1.7888,
+      "step": 15289
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6088922287951806e-05,
+      "loss": 1.9366,
+      "step": 15290
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6088446671515018e-05,
+      "loss": 2.3347,
+      "step": 15291
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6087971033191697e-05,
+      "loss": 1.7691,
+      "step": 15292
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.6087495372983566e-05,
+      "loss": 1.9502,
+      "step": 15293
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.608701969089233e-05,
+      "loss": 1.9286,
+      "step": 15294
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6086543986919693e-05,
+      "loss": 1.6795,
+      "step": 15295
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.6086068261067373e-05,
+      "loss": 1.884,
+      "step": 15296
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6085592513337073e-05,
+      "loss": 1.9645,
+      "step": 15297
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6085116743730513e-05,
+      "loss": 1.8537,
+      "step": 15298
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6084640952249396e-05,
+      "loss": 1.5992,
+      "step": 15299
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.608416513889543e-05,
+      "loss": 1.8691,
+      "step": 15300
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.608368930367033e-05,
+      "loss": 1.5934,
+      "step": 15301
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6083213446575808e-05,
+      "loss": 1.8546,
+      "step": 15302
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6082737567613572e-05,
+      "loss": 1.6751,
+      "step": 15303
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.608226166678533e-05,
+      "loss": 1.8843,
+      "step": 15304
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.60817857440928e-05,
+      "loss": 1.9777,
+      "step": 15305
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6081309799537688e-05,
+      "loss": 1.6137,
+      "step": 15306
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6080833833121702e-05,
+      "loss": 1.8645,
+      "step": 15307
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.608035784484656e-05,
+      "loss": 1.9677,
+      "step": 15308
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.6079881834713967e-05,
+      "loss": 2.1468,
+      "step": 15309
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6079405802725637e-05,
+      "loss": 2.0812,
+      "step": 15310
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6078929748883285e-05,
+      "loss": 1.7414,
+      "step": 15311
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6078453673188613e-05,
+      "loss": 1.5137,
+      "step": 15312
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6077977575643338e-05,
+      "loss": 2.0006,
+      "step": 15313
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6077501456249173e-05,
+      "loss": 2.2757,
+      "step": 15314
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6077025315007827e-05,
+      "loss": 1.727,
+      "step": 15315
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.6076549151921016e-05,
+      "loss": 1.7505,
+      "step": 15316
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6076072966990438e-05,
+      "loss": 1.9442,
+      "step": 15317
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6075596760217825e-05,
+      "loss": 1.6092,
+      "step": 15318
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6075120531604874e-05,
+      "loss": 1.9227,
+      "step": 15319
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6074644281153302e-05,
+      "loss": 1.7697,
+      "step": 15320
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6074168008864818e-05,
+      "loss": 1.7831,
+      "step": 15321
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.607369171474114e-05,
+      "loss": 1.9631,
+      "step": 15322
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6073215398783974e-05,
+      "loss": 1.8635,
+      "step": 15323
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.6072739060995035e-05,
+      "loss": 1.8506,
+      "step": 15324
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6072262701376038e-05,
+      "loss": 1.7675,
+      "step": 15325
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.607178631992869e-05,
+      "loss": 2.0028,
+      "step": 15326
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6071309916654706e-05,
+      "loss": 1.6289,
+      "step": 15327
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.60708334915558e-05,
+      "loss": 2.1983,
+      "step": 15328
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6070357044633683e-05,
+      "loss": 2.0574,
+      "step": 15329
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.606988057589007e-05,
+      "loss": 1.974,
+      "step": 15330
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.6069404085326667e-05,
+      "loss": 2.2107,
+      "step": 15331
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6068927572945193e-05,
+      "loss": 1.7446,
+      "step": 15332
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6068451038747363e-05,
+      "loss": 1.8004,
+      "step": 15333
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6067974482734886e-05,
+      "loss": 1.8555,
+      "step": 15334
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6067497904909472e-05,
+      "loss": 1.5225,
+      "step": 15335
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6067021305272844e-05,
+      "loss": 1.5027,
+      "step": 15336
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6066544683826703e-05,
+      "loss": 1.8859,
+      "step": 15337
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.606606804057277e-05,
+      "loss": 1.8348,
+      "step": 15338
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.606559137551276e-05,
+      "loss": 1.5448,
+      "step": 15339
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6065114688648382e-05,
+      "loss": 1.7683,
+      "step": 15340
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.606463797998135e-05,
+      "loss": 1.7548,
+      "step": 15341
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6064161249513383e-05,
+      "loss": 2.0451,
+      "step": 15342
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.6063684497246185e-05,
+      "loss": 1.8697,
+      "step": 15343
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6063207723181475e-05,
+      "loss": 1.9792,
+      "step": 15344
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.606273092732097e-05,
+      "loss": 1.9488,
+      "step": 15345
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.6062254109666383e-05,
+      "loss": 2.1153,
+      "step": 15346
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6061777270219424e-05,
+      "loss": 2.0844,
+      "step": 15347
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6061300408981807e-05,
+      "loss": 2.08,
+      "step": 15348
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.6060823525955253e-05,
+      "loss": 2.1677,
+      "step": 15349
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6060346621141466e-05,
+      "loss": 1.9996,
+      "step": 15350
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.605986969454217e-05,
+      "loss": 1.8912,
+      "step": 15351
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6059392746159074e-05,
+      "loss": 1.7018,
+      "step": 15352
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6058915775993894e-05,
+      "loss": 1.5331,
+      "step": 15353
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.605843878404835e-05,
+      "loss": 1.8468,
+      "step": 15354
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6057961770324144e-05,
+      "loss": 2.008,
+      "step": 15355
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6057484734823003e-05,
+      "loss": 1.8354,
+      "step": 15356
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6057007677546633e-05,
+      "loss": 1.8,
+      "step": 15357
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.6056530598496753e-05,
+      "loss": 1.6624,
+      "step": 15358
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.605605349767508e-05,
+      "loss": 1.7926,
+      "step": 15359
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.605557637508333e-05,
+      "loss": 1.4733,
+      "step": 15360
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6055099230723212e-05,
+      "loss": 2.0142,
+      "step": 15361
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.605462206459644e-05,
+      "loss": 1.6768,
+      "step": 15362
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.605414487670474e-05,
+      "loss": 1.8501,
+      "step": 15363
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6053667667049818e-05,
+      "loss": 1.8539,
+      "step": 15364
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6053190435633392e-05,
+      "loss": 1.8125,
+      "step": 15365
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.605271318245718e-05,
+      "loss": 1.9229,
+      "step": 15366
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.605223590752289e-05,
+      "loss": 1.7634,
+      "step": 15367
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.605175861083225e-05,
+      "loss": 1.8561,
+      "step": 15368
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.605128129238697e-05,
+      "loss": 1.8171,
+      "step": 15369
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.605080395218876e-05,
+      "loss": 1.5924,
+      "step": 15370
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.605032659023934e-05,
+      "loss": 2.0561,
+      "step": 15371
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.604984920654043e-05,
+      "loss": 1.8914,
+      "step": 15372
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6049371801093743e-05,
+      "loss": 2.0182,
+      "step": 15373
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6048894373900992e-05,
+      "loss": 2.0123,
+      "step": 15374
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6048416924963896e-05,
+      "loss": 1.853,
+      "step": 15375
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6047939454284172e-05,
+      "loss": 2.0438,
+      "step": 15376
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.604746196186354e-05,
+      "loss": 2.2792,
+      "step": 15377
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.604698444770371e-05,
+      "loss": 1.893,
+      "step": 15378
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6046506911806403e-05,
+      "loss": 1.9117,
+      "step": 15379
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.604602935417333e-05,
+      "loss": 2.0053,
+      "step": 15380
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.604555177480621e-05,
+      "loss": 2.2569,
+      "step": 15381
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6045074173706768e-05,
+      "loss": 1.8896,
+      "step": 15382
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6044596550876708e-05,
+      "loss": 1.9597,
+      "step": 15383
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.6044118906317753e-05,
+      "loss": 2.0149,
+      "step": 15384
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.6043641240031623e-05,
+      "loss": 1.5893,
+      "step": 15385
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.604316355202003e-05,
+      "loss": 1.9633,
+      "step": 15386
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.6042685842284694e-05,
+      "loss": 1.8533,
+      "step": 15387
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.6042208110827335e-05,
+      "loss": 1.7369,
+      "step": 15388
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.604173035764966e-05,
+      "loss": 1.676,
+      "step": 15389
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6041252582753396e-05,
+      "loss": 1.9629,
+      "step": 15390
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.604077478614026e-05,
+      "loss": 1.9797,
+      "step": 15391
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.6040296967811963e-05,
+      "loss": 2.0524,
+      "step": 15392
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.603981912777023e-05,
+      "loss": 1.9625,
+      "step": 15393
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6039341266016775e-05,
+      "loss": 1.9394,
+      "step": 15394
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.6038863382553315e-05,
+      "loss": 2.0401,
+      "step": 15395
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6038385477381573e-05,
+      "loss": 1.5233,
+      "step": 15396
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.603790755050326e-05,
+      "loss": 1.9259,
+      "step": 15397
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6037429601920097e-05,
+      "loss": 2.0763,
+      "step": 15398
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6036951631633802e-05,
+      "loss": 1.9082,
+      "step": 15399
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6036473639646096e-05,
+      "loss": 2.0092,
+      "step": 15400
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6035995625958693e-05,
+      "loss": 1.9121,
+      "step": 15401
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6035517590573315e-05,
+      "loss": 1.919,
+      "step": 15402
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6035039533491675e-05,
+      "loss": 1.7049,
+      "step": 15403
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.60345614547155e-05,
+      "loss": 1.7619,
+      "step": 15404
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6034083354246502e-05,
+      "loss": 2.1667,
+      "step": 15405
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.60336052320864e-05,
+      "loss": 1.9245,
+      "step": 15406
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6033127088236918e-05,
+      "loss": 1.8287,
+      "step": 15407
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.6032648922699763e-05,
+      "loss": 2.0119,
+      "step": 15408
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.603217073547667e-05,
+      "loss": 2.1093,
+      "step": 15409
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6031692526569344e-05,
+      "loss": 2.0351,
+      "step": 15410
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6031214295979514e-05,
+      "loss": 2.1711,
+      "step": 15411
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.6030736043708893e-05,
+      "loss": 1.7717,
+      "step": 15412
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6030257769759202e-05,
+      "loss": 2.0195,
+      "step": 15413
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.602977947413216e-05,
+      "loss": 1.7508,
+      "step": 15414
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6029301156829487e-05,
+      "loss": 1.875,
+      "step": 15415
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.6028822817852904e-05,
+      "loss": 2.3481,
+      "step": 15416
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.6028344457204127e-05,
+      "loss": 2.0849,
+      "step": 15417
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.6027866074884878e-05,
+      "loss": 1.8189,
+      "step": 15418
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6027387670896874e-05,
+      "loss": 1.4743,
+      "step": 15419
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.602690924524184e-05,
+      "loss": 1.839,
+      "step": 15420
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.602643079792149e-05,
+      "loss": 2.1197,
+      "step": 15421
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6025952328937548e-05,
+      "loss": 2.1715,
+      "step": 15422
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6025473838291734e-05,
+      "loss": 1.9969,
+      "step": 15423
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.6024995325985762e-05,
+      "loss": 2.0184,
+      "step": 15424
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6024516792021362e-05,
+      "loss": 1.9407,
+      "step": 15425
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6024038236400246e-05,
+      "loss": 1.9795,
+      "step": 15426
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.602355965912414e-05,
+      "loss": 1.6556,
+      "step": 15427
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.6023081060194756e-05,
+      "loss": 1.8192,
+      "step": 15428
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.602260243961382e-05,
+      "loss": 1.8391,
+      "step": 15429
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.602212379738306e-05,
+      "loss": 1.7213,
+      "step": 15430
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.6021645133504186e-05,
+      "loss": 1.6344,
+      "step": 15431
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6021166447978916e-05,
+      "loss": 1.7006,
+      "step": 15432
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6020687740808983e-05,
+      "loss": 1.9699,
+      "step": 15433
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.60202090119961e-05,
+      "loss": 1.6088,
+      "step": 15434
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.601973026154199e-05,
+      "loss": 1.7766,
+      "step": 15435
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.601925148944837e-05,
+      "loss": 1.7361,
+      "step": 15436
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.6018772695716966e-05,
+      "loss": 1.9963,
+      "step": 15437
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.60182938803495e-05,
+      "loss": 1.9347,
+      "step": 15438
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.601781504334769e-05,
+      "loss": 1.7551,
+      "step": 15439
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6017336184713256e-05,
+      "loss": 1.9637,
+      "step": 15440
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6016857304447922e-05,
+      "loss": 1.877,
+      "step": 15441
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.601637840255341e-05,
+      "loss": 1.7932,
+      "step": 15442
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.601589947903144e-05,
+      "loss": 1.7945,
+      "step": 15443
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6015420533883736e-05,
+      "loss": 1.9472,
+      "step": 15444
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.601494156711201e-05,
+      "loss": 1.6457,
+      "step": 15445
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6014462578718002e-05,
+      "loss": 1.9215,
+      "step": 15446
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.601398356870342e-05,
+      "loss": 1.8959,
+      "step": 15447
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6013504537069984e-05,
+      "loss": 1.9965,
+      "step": 15448
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6013025483819428e-05,
+      "loss": 2.0802,
+      "step": 15449
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6012546408953465e-05,
+      "loss": 2.1463,
+      "step": 15450
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6012067312473817e-05,
+      "loss": 1.8438,
+      "step": 15451
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.601158819438221e-05,
+      "loss": 1.9039,
+      "step": 15452
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.601110905468037e-05,
+      "loss": 2.0004,
+      "step": 15453
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.601062989337001e-05,
+      "loss": 2.0794,
+      "step": 15454
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.601015071045286e-05,
+      "loss": 1.8681,
+      "step": 15455
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.6009671505930636e-05,
+      "loss": 1.9219,
+      "step": 15456
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.6009192279805066e-05,
+      "loss": 1.8643,
+      "step": 15457
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6008713032077873e-05,
+      "loss": 1.8942,
+      "step": 15458
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.600823376275077e-05,
+      "loss": 1.5903,
+      "step": 15459
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6007754471825497e-05,
+      "loss": 1.6577,
+      "step": 15460
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.6007275159303763e-05,
+      "loss": 1.7506,
+      "step": 15461
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6006795825187294e-05,
+      "loss": 1.7748,
+      "step": 15462
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.600631646947782e-05,
+      "loss": 1.9786,
+      "step": 15463
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.6005837092177052e-05,
+      "loss": 1.6533,
+      "step": 15464
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6005357693286723e-05,
+      "loss": 2.1319,
+      "step": 15465
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.6004878272808553e-05,
+      "loss": 2.1053,
+      "step": 15466
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.6004398830744266e-05,
+      "loss": 1.8652,
+      "step": 15467
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6003919367095585e-05,
+      "loss": 1.925,
+      "step": 15468
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.6003439881864236e-05,
+      "loss": 1.9986,
+      "step": 15469
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6002960375051938e-05,
+      "loss": 1.6922,
+      "step": 15470
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.600248084666042e-05,
+      "loss": 1.8585,
+      "step": 15471
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6002001296691396e-05,
+      "loss": 1.8825,
+      "step": 15472
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.6001521725146604e-05,
+      "loss": 1.768,
+      "step": 15473
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6001042132027757e-05,
+      "loss": 2.0281,
+      "step": 15474
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.6000562517336584e-05,
+      "loss": 2.0577,
+      "step": 15475
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.6000082881074805e-05,
+      "loss": 2.4673,
+      "step": 15476
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5999603223244153e-05,
+      "loss": 1.7679,
+      "step": 15477
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5999123543846344e-05,
+      "loss": 1.7415,
+      "step": 15478
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5998643842883103e-05,
+      "loss": 2.0204,
+      "step": 15479
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5998164120356155e-05,
+      "loss": 1.8185,
+      "step": 15480
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5997684376267225e-05,
+      "loss": 1.9994,
+      "step": 15481
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5997204610618043e-05,
+      "loss": 2.0742,
+      "step": 15482
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5996724823410324e-05,
+      "loss": 1.8962,
+      "step": 15483
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.59962450146458e-05,
+      "loss": 1.8518,
+      "step": 15484
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.599576518432619e-05,
+      "loss": 1.5639,
+      "step": 15485
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5995285332453227e-05,
+      "loss": 1.9109,
+      "step": 15486
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.599480545902863e-05,
+      "loss": 1.662,
+      "step": 15487
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5994325564054122e-05,
+      "loss": 1.8342,
+      "step": 15488
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5993845647531432e-05,
+      "loss": 1.8761,
+      "step": 15489
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5993365709462287e-05,
+      "loss": 1.6534,
+      "step": 15490
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5992885749848408e-05,
+      "loss": 1.7612,
+      "step": 15491
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.599240576869152e-05,
+      "loss": 1.8418,
+      "step": 15492
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5991925765993355e-05,
+      "loss": 1.7729,
+      "step": 15493
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5991445741755633e-05,
+      "loss": 1.8442,
+      "step": 15494
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.599096569598008e-05,
+      "loss": 1.927,
+      "step": 15495
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.599048562866842e-05,
+      "loss": 1.7879,
+      "step": 15496
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5990005539822383e-05,
+      "loss": 1.9004,
+      "step": 15497
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.598952542944369e-05,
+      "loss": 1.776,
+      "step": 15498
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.598904529753407e-05,
+      "loss": 1.7587,
+      "step": 15499
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.598856514409525e-05,
+      "loss": 1.7736,
+      "step": 15500
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5988084969128955e-05,
+      "loss": 1.8519,
+      "step": 15501
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.598760477263691e-05,
+      "loss": 1.9906,
+      "step": 15502
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.598712455462084e-05,
+      "loss": 1.9803,
+      "step": 15503
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5986644315082478e-05,
+      "loss": 1.8634,
+      "step": 15504
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5986164054023542e-05,
+      "loss": 1.8369,
+      "step": 15505
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.598568377144576e-05,
+      "loss": 2.4751,
+      "step": 15506
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.598520346735086e-05,
+      "loss": 1.7862,
+      "step": 15507
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5984723141740578e-05,
+      "loss": 2.1489,
+      "step": 15508
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5984242794616624e-05,
+      "loss": 1.7883,
+      "step": 15509
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5983762425980733e-05,
+      "loss": 1.8943,
+      "step": 15510
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.598328203583463e-05,
+      "loss": 1.7004,
+      "step": 15511
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5982801624180048e-05,
+      "loss": 1.9306,
+      "step": 15512
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5982321191018705e-05,
+      "loss": 2.0549,
+      "step": 15513
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5981840736352334e-05,
+      "loss": 1.9021,
+      "step": 15514
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.598136026018266e-05,
+      "loss": 2.2541,
+      "step": 15515
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.598087976251141e-05,
+      "loss": 2.0006,
+      "step": 15516
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.5980399243340312e-05,
+      "loss": 2.3221,
+      "step": 15517
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.597991870267109e-05,
+      "loss": 1.9989,
+      "step": 15518
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.597943814050548e-05,
+      "loss": 2.0933,
+      "step": 15519
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.59789575568452e-05,
+      "loss": 2.0512,
+      "step": 15520
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.597847695169198e-05,
+      "loss": 2.0598,
+      "step": 15521
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5977996325047556e-05,
+      "loss": 1.7689,
+      "step": 15522
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5977515676913642e-05,
+      "loss": 1.8068,
+      "step": 15523
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5977035007291977e-05,
+      "loss": 1.8915,
+      "step": 15524
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5976554316184283e-05,
+      "loss": 1.7736,
+      "step": 15525
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5976073603592287e-05,
+      "loss": 1.898,
+      "step": 15526
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5975592869517726e-05,
+      "loss": 1.8053,
+      "step": 15527
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.597511211396232e-05,
+      "loss": 1.5147,
+      "step": 15528
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5974631336927795e-05,
+      "loss": 2.1759,
+      "step": 15529
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5974150538415883e-05,
+      "loss": 2.024,
+      "step": 15530
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.597366971842832e-05,
+      "loss": 1.7214,
+      "step": 15531
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.597318887696682e-05,
+      "loss": 1.7395,
+      "step": 15532
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5972708014033122e-05,
+      "loss": 1.7713,
+      "step": 15533
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.597222712962895e-05,
+      "loss": 1.9356,
+      "step": 15534
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5971746223756036e-05,
+      "loss": 1.817,
+      "step": 15535
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5971265296416105e-05,
+      "loss": 1.6686,
+      "step": 15536
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5970784347610886e-05,
+      "loss": 1.925,
+      "step": 15537
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5970303377342112e-05,
+      "loss": 1.778,
+      "step": 15538
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.5969822385611506e-05,
+      "loss": 1.9492,
+      "step": 15539
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5969341372420802e-05,
+      "loss": 1.7844,
+      "step": 15540
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5968860337771728e-05,
+      "loss": 1.908,
+      "step": 15541
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5968379281666014e-05,
+      "loss": 1.7387,
+      "step": 15542
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5967898204105387e-05,
+      "loss": 2.2933,
+      "step": 15543
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5967417105091576e-05,
+      "loss": 1.8526,
+      "step": 15544
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5966935984626314e-05,
+      "loss": 1.9595,
+      "step": 15545
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5966454842711322e-05,
+      "loss": 2.2096,
+      "step": 15546
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5965973679348344e-05,
+      "loss": 2.0681,
+      "step": 15547
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5965492494539095e-05,
+      "loss": 1.7269,
+      "step": 15548
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5965011288285316e-05,
+      "loss": 2.0128,
+      "step": 15549
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.596453006058873e-05,
+      "loss": 2.2359,
+      "step": 15550
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5964048811451068e-05,
+      "loss": 1.91,
+      "step": 15551
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5963567540874065e-05,
+      "loss": 2.091,
+      "step": 15552
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5963086248859443e-05,
+      "loss": 1.8448,
+      "step": 15553
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.596260493540894e-05,
+      "loss": 2.1645,
+      "step": 15554
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.596212360052428e-05,
+      "loss": 1.673,
+      "step": 15555
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5961642244207193e-05,
+      "loss": 1.8239,
+      "step": 15556
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5961160866459417e-05,
+      "loss": 1.8945,
+      "step": 15557
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5960679467282677e-05,
+      "loss": 2.0108,
+      "step": 15558
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.59601980466787e-05,
+      "loss": 1.6736,
+      "step": 15559
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.595971660464922e-05,
+      "loss": 1.9793,
+      "step": 15560
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5959235141195975e-05,
+      "loss": 1.8058,
+      "step": 15561
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5958753656320682e-05,
+      "loss": 1.6804,
+      "step": 15562
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5958272150025082e-05,
+      "loss": 1.7634,
+      "step": 15563
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5957790622310906e-05,
+      "loss": 1.6932,
+      "step": 15564
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5957309073179877e-05,
+      "loss": 1.958,
+      "step": 15565
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.595682750263373e-05,
+      "loss": 1.8418,
+      "step": 15566
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.59563459106742e-05,
+      "loss": 1.5765,
+      "step": 15567
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5955864297303014e-05,
+      "loss": 1.6078,
+      "step": 15568
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5955382662521904e-05,
+      "loss": 2.0039,
+      "step": 15569
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5954901006332603e-05,
+      "loss": 2.0179,
+      "step": 15570
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5954419328736837e-05,
+      "loss": 1.8493,
+      "step": 15571
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5953937629736346e-05,
+      "loss": 1.7467,
+      "step": 15572
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5953455909332857e-05,
+      "loss": 1.7184,
+      "step": 15573
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.59529741675281e-05,
+      "loss": 1.8811,
+      "step": 15574
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5952492404323813e-05,
+      "loss": 1.8829,
+      "step": 15575
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5952010619721718e-05,
+      "loss": 1.6797,
+      "step": 15576
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5951528813723554e-05,
+      "loss": 1.59,
+      "step": 15577
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5951046986331052e-05,
+      "loss": 1.5714,
+      "step": 15578
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5950565137545947e-05,
+      "loss": 1.8984,
+      "step": 15579
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5950083267369963e-05,
+      "loss": 1.5342,
+      "step": 15580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.594960137580484e-05,
+      "loss": 1.6077,
+      "step": 15581
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5949119462852302e-05,
+      "loss": 2.367,
+      "step": 15582
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.594863752851409e-05,
+      "loss": 1.5944,
+      "step": 15583
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5948155572791934e-05,
+      "loss": 1.9621,
+      "step": 15584
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5947673595687564e-05,
+      "loss": 1.8336,
+      "step": 15585
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5947191597202715e-05,
+      "loss": 1.7855,
+      "step": 15586
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5946709577339116e-05,
+      "loss": 2.2103,
+      "step": 15587
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5946227536098503e-05,
+      "loss": 1.7952,
+      "step": 15588
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5945745473482607e-05,
+      "loss": 1.648,
+      "step": 15589
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5945263389493166e-05,
+      "loss": 2.0692,
+      "step": 15590
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5944781284131908e-05,
+      "loss": 2.3094,
+      "step": 15591
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5944299157400568e-05,
+      "loss": 2.0025,
+      "step": 15592
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5943817009300877e-05,
+      "loss": 1.7671,
+      "step": 15593
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5943334839834566e-05,
+      "loss": 1.849,
+      "step": 15594
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.594285264900338e-05,
+      "loss": 1.9757,
+      "step": 15595
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5942370436809036e-05,
+      "loss": 1.8151,
+      "step": 15596
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.594188820325328e-05,
+      "loss": 1.802,
+      "step": 15597
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.594140594833784e-05,
+      "loss": 2.0581,
+      "step": 15598
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5940923672064453e-05,
+      "loss": 1.7837,
+      "step": 15599
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5940441374434847e-05,
+      "loss": 1.7864,
+      "step": 15600
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5939959055450762e-05,
+      "loss": 1.7266,
+      "step": 15601
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.593947671511393e-05,
+      "loss": 2.1685,
+      "step": 15602
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5938994353426077e-05,
+      "loss": 2.0639,
+      "step": 15603
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5938511970388947e-05,
+      "loss": 1.8211,
+      "step": 15604
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5938029566004276e-05,
+      "loss": 1.6482,
+      "step": 15605
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.593754714027379e-05,
+      "loss": 1.621,
+      "step": 15606
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5937064693199223e-05,
+      "loss": 2.4195,
+      "step": 15607
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.5936582224782317e-05,
+      "loss": 1.4403,
+      "step": 15608
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5936099735024795e-05,
+      "loss": 2.1022,
+      "step": 15609
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5935617223928406e-05,
+      "loss": 1.8182,
+      "step": 15610
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5935134691494875e-05,
+      "loss": 2.0771,
+      "step": 15611
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5934652137725938e-05,
+      "loss": 1.9319,
+      "step": 15612
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.593416956262333e-05,
+      "loss": 1.8101,
+      "step": 15613
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5933686966188783e-05,
+      "loss": 1.846,
+      "step": 15614
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5933204348424038e-05,
+      "loss": 2.1138,
+      "step": 15615
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5932721709330823e-05,
+      "loss": 1.65,
+      "step": 15616
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5932239048910877e-05,
+      "loss": 1.8298,
+      "step": 15617
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5931756367165938e-05,
+      "loss": 1.862,
+      "step": 15618
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5931273664097738e-05,
+      "loss": 2.0227,
+      "step": 15619
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5930790939708007e-05,
+      "loss": 2.0659,
+      "step": 15620
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.593030819399849e-05,
+      "loss": 1.8463,
+      "step": 15621
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5929825426970914e-05,
+      "loss": 1.6084,
+      "step": 15622
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.592934263862702e-05,
+      "loss": 1.813,
+      "step": 15623
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.592885982896854e-05,
+      "loss": 1.9907,
+      "step": 15624
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5928376997997213e-05,
+      "loss": 2.0137,
+      "step": 15625
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.592789414571477e-05,
+      "loss": 1.9642,
+      "step": 15626
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5927411272122954e-05,
+      "loss": 1.8366,
+      "step": 15627
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5926928377223495e-05,
+      "loss": 2.0121,
+      "step": 15628
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.5926445461018126e-05,
+      "loss": 1.6157,
+      "step": 15629
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.592596252350859e-05,
+      "loss": 2.0121,
+      "step": 15630
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.592547956469662e-05,
+      "loss": 1.7798,
+      "step": 15631
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5924996584583952e-05,
+      "loss": 1.8154,
+      "step": 15632
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.592451358317232e-05,
+      "loss": 1.6556,
+      "step": 15633
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5924030560463467e-05,
+      "loss": 1.6486,
+      "step": 15634
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5923547516459125e-05,
+      "loss": 1.7221,
+      "step": 15635
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.592306445116103e-05,
+      "loss": 1.6814,
+      "step": 15636
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5922581364570916e-05,
+      "loss": 1.8126,
+      "step": 15637
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.592209825669053e-05,
+      "loss": 2.4742,
+      "step": 15638
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5921615127521594e-05,
+      "loss": 1.7815,
+      "step": 15639
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5921131977065854e-05,
+      "loss": 1.9632,
+      "step": 15640
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5920648805325044e-05,
+      "loss": 1.4875,
+      "step": 15641
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.59201656123009e-05,
+      "loss": 1.5124,
+      "step": 15642
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5919682397995168e-05,
+      "loss": 2.4174,
+      "step": 15643
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.591919916240957e-05,
+      "loss": 1.693,
+      "step": 15644
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5918715905545856e-05,
+      "loss": 1.6922,
+      "step": 15645
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5918232627405754e-05,
+      "loss": 1.8378,
+      "step": 15646
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5917749327991007e-05,
+      "loss": 1.666,
+      "step": 15647
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.5917266007303353e-05,
+      "loss": 1.6796,
+      "step": 15648
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5916782665344526e-05,
+      "loss": 1.7405,
+      "step": 15649
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5916299302116262e-05,
+      "loss": 1.6335,
+      "step": 15650
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5915815917620304e-05,
+      "loss": 2.0905,
+      "step": 15651
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5915332511858386e-05,
+      "loss": 1.8444,
+      "step": 15652
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5914849084832247e-05,
+      "loss": 1.6067,
+      "step": 15653
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5914365636543622e-05,
+      "loss": 1.982,
+      "step": 15654
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5913882166994254e-05,
+      "loss": 1.8898,
+      "step": 15655
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5913398676185877e-05,
+      "loss": 1.7589,
+      "step": 15656
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5912915164120232e-05,
+      "loss": 2.0831,
+      "step": 15657
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5912431630799055e-05,
+      "loss": 2.0333,
+      "step": 15658
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5911948076224082e-05,
+      "loss": 1.8463,
+      "step": 15659
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5911464500397053e-05,
+      "loss": 1.8801,
+      "step": 15660
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.591098090331971e-05,
+      "loss": 1.6779,
+      "step": 15661
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.591049728499379e-05,
+      "loss": 2.2716,
+      "step": 15662
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5910013645421027e-05,
+      "loss": 2.0137,
+      "step": 15663
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5909529984603166e-05,
+      "loss": 1.8466,
+      "step": 15664
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.590904630254194e-05,
+      "loss": 1.7653,
+      "step": 15665
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5908562599239088e-05,
+      "loss": 2.0341,
+      "step": 15666
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.590807887469635e-05,
+      "loss": 2.2386,
+      "step": 15667
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5907595128915473e-05,
+      "loss": 1.8649,
+      "step": 15668
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5907111361898183e-05,
+      "loss": 1.9406,
+      "step": 15669
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.590662757364622e-05,
+      "loss": 1.5103,
+      "step": 15670
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.590614376416134e-05,
+      "loss": 1.8343,
+      "step": 15671
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.590565993344526e-05,
+      "loss": 1.6495,
+      "step": 15672
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5905176081499734e-05,
+      "loss": 1.8885,
+      "step": 15673
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5904692208326495e-05,
+      "loss": 1.9569,
+      "step": 15674
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5904208313927284e-05,
+      "loss": 1.5457,
+      "step": 15675
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.590372439830384e-05,
+      "loss": 1.8176,
+      "step": 15676
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.59032404614579e-05,
+      "loss": 1.8959,
+      "step": 15677
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5902756503391216e-05,
+      "loss": 2.005,
+      "step": 15678
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5902272524105507e-05,
+      "loss": 1.8539,
+      "step": 15679
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5901788523602532e-05,
+      "loss": 1.9882,
+      "step": 15680
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5901304501884017e-05,
+      "loss": 2.2225,
+      "step": 15681
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5900820458951714e-05,
+      "loss": 1.9674,
+      "step": 15682
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5900336394807355e-05,
+      "loss": 2.1083,
+      "step": 15683
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5899852309452678e-05,
+      "loss": 2.1376,
+      "step": 15684
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5899368202889434e-05,
+      "loss": 1.8989,
+      "step": 15685
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.589888407511935e-05,
+      "loss": 2.1158,
+      "step": 15686
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5898399926144175e-05,
+      "loss": 2.2114,
+      "step": 15687
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.589791575596565e-05,
+      "loss": 1.7245,
+      "step": 15688
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.589743156458551e-05,
+      "loss": 2.0835,
+      "step": 15689
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.58969473520055e-05,
+      "loss": 2.0469,
+      "step": 15690
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.589646311822736e-05,
+      "loss": 1.8027,
+      "step": 15691
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5895978863252827e-05,
+      "loss": 1.7922,
+      "step": 15692
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5895494587083642e-05,
+      "loss": 2.3471,
+      "step": 15693
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.589501028972155e-05,
+      "loss": 1.7964,
+      "step": 15694
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5894525971168292e-05,
+      "loss": 1.8883,
+      "step": 15695
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5894041631425606e-05,
+      "loss": 2.3108,
+      "step": 15696
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5893557270495237e-05,
+      "loss": 1.8688,
+      "step": 15697
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5893072888378918e-05,
+      "loss": 1.6939,
+      "step": 15698
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.58925884850784e-05,
+      "loss": 1.7437,
+      "step": 15699
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5892104060595416e-05,
+      "loss": 1.6541,
+      "step": 15700
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5891619614931714e-05,
+      "loss": 2.1759,
+      "step": 15701
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5891135148089034e-05,
+      "loss": 1.7253,
+      "step": 15702
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5890650660069113e-05,
+      "loss": 1.8045,
+      "step": 15703
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.58901661508737e-05,
+      "loss": 1.9004,
+      "step": 15704
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.5889681620504532e-05,
+      "loss": 2.0525,
+      "step": 15705
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5889197068963348e-05,
+      "loss": 1.9096,
+      "step": 15706
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5888712496251896e-05,
+      "loss": 1.9913,
+      "step": 15707
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5888227902371912e-05,
+      "loss": 1.8232,
+      "step": 15708
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5887743287325144e-05,
+      "loss": 1.5427,
+      "step": 15709
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.588725865111333e-05,
+      "loss": 2.1097,
+      "step": 15710
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5886773993738218e-05,
+      "loss": 2.1256,
+      "step": 15711
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5886289315201543e-05,
+      "loss": 1.751,
+      "step": 15712
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5885804615505047e-05,
+      "loss": 1.7613,
+      "step": 15713
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5885319894650477e-05,
+      "loss": 1.7358,
+      "step": 15714
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5884835152639575e-05,
+      "loss": 1.8126,
+      "step": 15715
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.588435038947408e-05,
+      "loss": 2.0031,
+      "step": 15716
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.588386560515574e-05,
+      "loss": 2.337,
+      "step": 15717
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5883380799686295e-05,
+      "loss": 2.0544,
+      "step": 15718
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5882895973067485e-05,
+      "loss": 1.8433,
+      "step": 15719
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5882411125301054e-05,
+      "loss": 1.9657,
+      "step": 15720
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5881926256388747e-05,
+      "loss": 1.7223,
+      "step": 15721
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.588144136633231e-05,
+      "loss": 1.8855,
+      "step": 15722
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5880956455133478e-05,
+      "loss": 1.9657,
+      "step": 15723
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.5880471522794002e-05,
+      "loss": 1.9709,
+      "step": 15724
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5879986569315617e-05,
+      "loss": 2.0199,
+      "step": 15725
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5879501594700073e-05,
+      "loss": 1.9355,
+      "step": 15726
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.587901659894911e-05,
+      "loss": 1.8305,
+      "step": 15727
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5878531582064475e-05,
+      "loss": 1.7237,
+      "step": 15728
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5878046544047912e-05,
+      "loss": 1.8519,
+      "step": 15729
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5877561484901157e-05,
+      "loss": 1.8889,
+      "step": 15730
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.587707640462596e-05,
+      "loss": 1.7081,
+      "step": 15731
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.587659130322406e-05,
+      "loss": 1.7853,
+      "step": 15732
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.587610618069721e-05,
+      "loss": 1.8036,
+      "step": 15733
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5875621037047143e-05,
+      "loss": 1.9898,
+      "step": 15734
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.587513587227561e-05,
+      "loss": 2.1139,
+      "step": 15735
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5874650686384352e-05,
+      "loss": 2.2048,
+      "step": 15736
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5874165479375114e-05,
+      "loss": 1.9878,
+      "step": 15737
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5873680251249642e-05,
+      "loss": 2.0272,
+      "step": 15738
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.587319500200968e-05,
+      "loss": 1.9013,
+      "step": 15739
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5872709731656966e-05,
+      "loss": 1.8785,
+      "step": 15740
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.587222444019325e-05,
+      "loss": 2.014,
+      "step": 15741
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.587173912762028e-05,
+      "loss": 1.6843,
+      "step": 15742
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5871253793939794e-05,
+      "loss": 1.8082,
+      "step": 15743
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.587076843915354e-05,
+      "loss": 1.9786,
+      "step": 15744
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5870283063263257e-05,
+      "loss": 1.9527,
+      "step": 15745
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5869797666270702e-05,
+      "loss": 1.7026,
+      "step": 15746
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.5869312248177608e-05,
+      "loss": 1.9339,
+      "step": 15747
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5868826808985726e-05,
+      "loss": 1.7851,
+      "step": 15748
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.375,
+      "learning_rate": 1.58683413486968e-05,
+      "loss": 2.1855,
+      "step": 15749
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5867855867312574e-05,
+      "loss": 1.8766,
+      "step": 15750
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5867370364834795e-05,
+      "loss": 1.7563,
+      "step": 15751
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5866884841265207e-05,
+      "loss": 1.7628,
+      "step": 15752
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5866399296605553e-05,
+      "loss": 1.9039,
+      "step": 15753
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5865913730857583e-05,
+      "loss": 1.6092,
+      "step": 15754
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.586542814402304e-05,
+      "loss": 1.9869,
+      "step": 15755
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5864942536103672e-05,
+      "loss": 1.997,
+      "step": 15756
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.586445690710122e-05,
+      "loss": 1.7904,
+      "step": 15757
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.5863971257017433e-05,
+      "loss": 2.0289,
+      "step": 15758
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5863485585854055e-05,
+      "loss": 1.6298,
+      "step": 15759
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5862999893612835e-05,
+      "loss": 1.6442,
+      "step": 15760
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5862514180295514e-05,
+      "loss": 1.6794,
+      "step": 15761
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5862028445903844e-05,
+      "loss": 1.969,
+      "step": 15762
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5861542690439566e-05,
+      "loss": 1.6848,
+      "step": 15763
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.586105691390443e-05,
+      "loss": 1.5926,
+      "step": 15764
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5860571116300178e-05,
+      "loss": 2.0105,
+      "step": 15765
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5860085297628563e-05,
+      "loss": 1.7263,
+      "step": 15766
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5859599457891324e-05,
+      "loss": 1.9783,
+      "step": 15767
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.5859113597090212e-05,
+      "loss": 3.2235,
+      "step": 15768
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5858627715226972e-05,
+      "loss": 1.803,
+      "step": 15769
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5858141812303348e-05,
+      "loss": 2.1835,
+      "step": 15770
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5857655888321093e-05,
+      "loss": 1.9827,
+      "step": 15771
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5857169943281948e-05,
+      "loss": 1.888,
+      "step": 15772
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5856683977187664e-05,
+      "loss": 1.8642,
+      "step": 15773
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5856197990039987e-05,
+      "loss": 1.8896,
+      "step": 15774
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.585571198184066e-05,
+      "loss": 1.8496,
+      "step": 15775
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.585522595259144e-05,
+      "loss": 1.6796,
+      "step": 15776
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.585473990229406e-05,
+      "loss": 1.6717,
+      "step": 15777
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5854253830950278e-05,
+      "loss": 2.0084,
+      "step": 15778
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5853767738561837e-05,
+      "loss": 1.6373,
+      "step": 15779
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5853281625130485e-05,
+      "loss": 1.9493,
+      "step": 15780
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.585279549065797e-05,
+      "loss": 2.2044,
+      "step": 15781
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.375,
+      "learning_rate": 1.585230933514604e-05,
+      "loss": 2.032,
+      "step": 15782
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.5851823158596446e-05,
+      "loss": 1.9108,
+      "step": 15783
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5851336961010927e-05,
+      "loss": 2.0267,
+      "step": 15784
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5850850742391233e-05,
+      "loss": 1.6993,
+      "step": 15785
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.585036450273912e-05,
+      "loss": 1.7295,
+      "step": 15786
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5849878242056326e-05,
+      "loss": 1.989,
+      "step": 15787
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5849391960344607e-05,
+      "loss": 1.7938,
+      "step": 15788
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5848905657605704e-05,
+      "loss": 1.7394,
+      "step": 15789
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.5848419333841367e-05,
+      "loss": 1.5863,
+      "step": 15790
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.584793298905335e-05,
+      "loss": 1.6575,
+      "step": 15791
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5847446623243393e-05,
+      "loss": 2.1545,
+      "step": 15792
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.584696023641325e-05,
+      "loss": 2.3014,
+      "step": 15793
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.584647382856467e-05,
+      "loss": 2.0766,
+      "step": 15794
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5845987399699396e-05,
+      "loss": 2.0536,
+      "step": 15795
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5845500949819182e-05,
+      "loss": 2.0592,
+      "step": 15796
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5845014478925774e-05,
+      "loss": 1.6281,
+      "step": 15797
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5844527987020918e-05,
+      "loss": 2.151,
+      "step": 15798
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5844041474106372e-05,
+      "loss": 2.1714,
+      "step": 15799
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.584355494018387e-05,
+      "loss": 1.7302,
+      "step": 15800
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.584306838525518e-05,
+      "loss": 2.0814,
+      "step": 15801
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5842581809322035e-05,
+      "loss": 1.9816,
+      "step": 15802
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.584209521238619e-05,
+      "loss": 1.6815,
+      "step": 15803
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.58416085944494e-05,
+      "loss": 1.8968,
+      "step": 15804
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5841121955513403e-05,
+      "loss": 1.7326,
+      "step": 15805
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5840635295579954e-05,
+      "loss": 1.9638,
+      "step": 15806
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5840148614650806e-05,
+      "loss": 1.6356,
+      "step": 15807
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5839661912727705e-05,
+      "loss": 2.1133,
+      "step": 15808
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5839175189812393e-05,
+      "loss": 1.8106,
+      "step": 15809
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5838688445906636e-05,
+      "loss": 1.802,
+      "step": 15810
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5838201681012172e-05,
+      "loss": 1.9009,
+      "step": 15811
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5837714895130752e-05,
+      "loss": 2.0391,
+      "step": 15812
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5837228088264128e-05,
+      "loss": 1.9156,
+      "step": 15813
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.583674126041405e-05,
+      "loss": 1.8105,
+      "step": 15814
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5836254411582268e-05,
+      "loss": 1.7934,
+      "step": 15815
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.583576754177053e-05,
+      "loss": 1.8044,
+      "step": 15816
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.583528065098059e-05,
+      "loss": 1.9199,
+      "step": 15817
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5834793739214195e-05,
+      "loss": 1.7883,
+      "step": 15818
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5834306806473098e-05,
+      "loss": 2.0143,
+      "step": 15819
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.5833819852759045e-05,
+      "loss": 2.0449,
+      "step": 15820
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.583333287807379e-05,
+      "loss": 2.0592,
+      "step": 15821
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5832845882419082e-05,
+      "loss": 1.6136,
+      "step": 15822
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5832358865796675e-05,
+      "loss": 1.624,
+      "step": 15823
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5831871828208315e-05,
+      "loss": 1.9294,
+      "step": 15824
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5831384769655758e-05,
+      "loss": 2.1877,
+      "step": 15825
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.583089769014075e-05,
+      "loss": 1.6624,
+      "step": 15826
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5830410589665043e-05,
+      "loss": 1.9583,
+      "step": 15827
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.582992346823039e-05,
+      "loss": 1.7834,
+      "step": 15828
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.582943632583854e-05,
+      "loss": 1.648,
+      "step": 15829
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5828949162491244e-05,
+      "loss": 1.6399,
+      "step": 15830
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5828461978190256e-05,
+      "loss": 2.0747,
+      "step": 15831
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5827974772937324e-05,
+      "loss": 1.8593,
+      "step": 15832
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5827487546734202e-05,
+      "loss": 2.0109,
+      "step": 15833
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.582700029958264e-05,
+      "loss": 1.7803,
+      "step": 15834
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.582651303148439e-05,
+      "loss": 1.5833,
+      "step": 15835
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.582602574244121e-05,
+      "loss": 1.4555,
+      "step": 15836
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5825538432454836e-05,
+      "loss": 1.7657,
+      "step": 15837
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5825051101527034e-05,
+      "loss": 1.9493,
+      "step": 15838
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.375,
+      "learning_rate": 1.582456374965955e-05,
+      "loss": 3.3235,
+      "step": 15839
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5824076376854134e-05,
+      "loss": 1.5342,
+      "step": 15840
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5823588983112544e-05,
+      "loss": 1.6928,
+      "step": 15841
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.5823101568436527e-05,
+      "loss": 1.9,
+      "step": 15842
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.582261413282784e-05,
+      "loss": 1.9766,
+      "step": 15843
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.582212667628823e-05,
+      "loss": 1.9601,
+      "step": 15844
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.582163919881945e-05,
+      "loss": 1.8216,
+      "step": 15845
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5821151700423256e-05,
+      "loss": 1.7235,
+      "step": 15846
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.5820664181101398e-05,
+      "loss": 1.8068,
+      "step": 15847
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5820176640855627e-05,
+      "loss": 1.9517,
+      "step": 15848
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5819689079687698e-05,
+      "loss": 1.7475,
+      "step": 15849
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5819201497599363e-05,
+      "loss": 1.6656,
+      "step": 15850
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5818713894592377e-05,
+      "loss": 1.6985,
+      "step": 15851
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.875,
+      "learning_rate": 1.581822627066849e-05,
+      "loss": 1.8697,
+      "step": 15852
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5817738625829456e-05,
+      "loss": 2.0614,
+      "step": 15853
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5817250960077024e-05,
+      "loss": 1.9743,
+      "step": 15854
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5816763273412955e-05,
+      "loss": 2.171,
+      "step": 15855
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5816275565838993e-05,
+      "loss": 1.9422,
+      "step": 15856
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5815787837356897e-05,
+      "loss": 2.0558,
+      "step": 15857
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5815300087968425e-05,
+      "loss": 1.9826,
+      "step": 15858
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5814812317675315e-05,
+      "loss": 1.6698,
+      "step": 15859
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.581432452647934e-05,
+      "loss": 1.7637,
+      "step": 15860
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5813836714382237e-05,
+      "loss": 2.1321,
+      "step": 15861
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5813348881385766e-05,
+      "loss": 2.2052,
+      "step": 15862
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5812861027491683e-05,
+      "loss": 2.0747,
+      "step": 15863
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5812373152701735e-05,
+      "loss": 1.842,
+      "step": 15864
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5811885257017686e-05,
+      "loss": 1.7789,
+      "step": 15865
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.581139734044128e-05,
+      "loss": 2.2857,
+      "step": 15866
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5810909402974277e-05,
+      "loss": 1.7152,
+      "step": 15867
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5810421444618427e-05,
+      "loss": 2.2024,
+      "step": 15868
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5809933465375486e-05,
+      "loss": 2.0141,
+      "step": 15869
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.580944546524721e-05,
+      "loss": 2.2057,
+      "step": 15870
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.375,
+      "learning_rate": 1.580895744423535e-05,
+      "loss": 2.0936,
+      "step": 15871
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5808469402341664e-05,
+      "loss": 1.9788,
+      "step": 15872
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.5807981339567905e-05,
+      "loss": 2.1053,
+      "step": 15873
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.580749325591582e-05,
+      "loss": 1.5828,
+      "step": 15874
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5807005151387176e-05,
+      "loss": 1.6651,
+      "step": 15875
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.580651702598372e-05,
+      "loss": 1.6616,
+      "step": 15876
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.580602887970721e-05,
+      "loss": 2.0163,
+      "step": 15877
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.58055407125594e-05,
+      "loss": 1.789,
+      "step": 15878
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.580505252454204e-05,
+      "loss": 1.7858,
+      "step": 15879
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.580456431565689e-05,
+      "loss": 2.0457,
+      "step": 15880
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5804076085905706e-05,
+      "loss": 1.9072,
+      "step": 15881
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5803587835290238e-05,
+      "loss": 2.0512,
+      "step": 15882
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5803099563812247e-05,
+      "loss": 1.8453,
+      "step": 15883
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5802611271473484e-05,
+      "loss": 1.6482,
+      "step": 15884
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0,
+      "learning_rate": 1.580212295827571e-05,
+      "loss": 2.481,
+      "step": 15885
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.580163462422067e-05,
+      "loss": 1.9041,
+      "step": 15886
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.375,
+      "learning_rate": 1.580114626931013e-05,
+      "loss": 1.9389,
+      "step": 15887
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.580065789354584e-05,
+      "loss": 1.5936,
+      "step": 15888
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5800169496929555e-05,
+      "loss": 2.0889,
+      "step": 15889
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5799681079463032e-05,
+      "loss": 1.7527,
+      "step": 15890
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5799192641148028e-05,
+      "loss": 1.6639,
+      "step": 15891
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5798704181986302e-05,
+      "loss": 1.9465,
+      "step": 15892
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.57982157019796e-05,
+      "loss": 1.6049,
+      "step": 15893
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5797727201129688e-05,
+      "loss": 1.854,
+      "step": 15894
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5797238679438316e-05,
+      "loss": 1.989,
+      "step": 15895
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.579675013690724e-05,
+      "loss": 2.042,
+      "step": 15896
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5796261573538223e-05,
+      "loss": 1.4893,
+      "step": 15897
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5795772989333014e-05,
+      "loss": 1.979,
+      "step": 15898
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.579528438429337e-05,
+      "loss": 1.8271,
+      "step": 15899
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.579479575842105e-05,
+      "loss": 2.1034,
+      "step": 15900
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.579430711171781e-05,
+      "loss": 1.7195,
+      "step": 15901
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5793818444185406e-05,
+      "loss": 1.9027,
+      "step": 15902
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5793329755825594e-05,
+      "loss": 1.7341,
+      "step": 15903
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5792841046640136e-05,
+      "loss": 1.979,
+      "step": 15904
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.579235231663078e-05,
+      "loss": 1.8111,
+      "step": 15905
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.579186356579929e-05,
+      "loss": 1.5878,
+      "step": 15906
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5791374794147417e-05,
+      "loss": 1.7382,
+      "step": 15907
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5790886001676925e-05,
+      "loss": 1.898,
+      "step": 15908
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5790397188389562e-05,
+      "loss": 1.9517,
+      "step": 15909
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5789908354287097e-05,
+      "loss": 1.8748,
+      "step": 15910
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5789419499371276e-05,
+      "loss": 1.8879,
+      "step": 15911
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5788930623643863e-05,
+      "loss": 1.695,
+      "step": 15912
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5788441727106612e-05,
+      "loss": 1.7542,
+      "step": 15913
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5787952809761286e-05,
+      "loss": 1.7285,
+      "step": 15914
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5787463871609635e-05,
+      "loss": 2.2094,
+      "step": 15915
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.578697491265342e-05,
+      "loss": 1.5126,
+      "step": 15916
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.57864859328944e-05,
+      "loss": 1.8088,
+      "step": 15917
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.5785996932334332e-05,
+      "loss": 1.8419,
+      "step": 15918
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.578550791097497e-05,
+      "loss": 2.1285,
+      "step": 15919
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5785018868818077e-05,
+      "loss": 2.0245,
+      "step": 15920
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5784529805865412e-05,
+      "loss": 1.6316,
+      "step": 15921
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5784040722118724e-05,
+      "loss": 1.7541,
+      "step": 15922
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5783551617579784e-05,
+      "loss": 2.0313,
+      "step": 15923
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5783062492250342e-05,
+      "loss": 1.6881,
+      "step": 15924
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5782573346132157e-05,
+      "loss": 1.9534,
+      "step": 15925
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.578208417922699e-05,
+      "loss": 2.2361,
+      "step": 15926
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5781594991536597e-05,
+      "loss": 1.8413,
+      "step": 15927
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5781105783062736e-05,
+      "loss": 1.8867,
+      "step": 15928
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.578061655380717e-05,
+      "loss": 1.709,
+      "step": 15929
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.578012730377165e-05,
+      "loss": 1.907,
+      "step": 15930
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.577963803295794e-05,
+      "loss": 2.2975,
+      "step": 15931
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5779148741367804e-05,
+      "loss": 1.7382,
+      "step": 15932
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5778659429002987e-05,
+      "loss": 2.0361,
+      "step": 15933
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.577817009586526e-05,
+      "loss": 1.8734,
+      "step": 15934
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.577768074195638e-05,
+      "loss": 1.989,
+      "step": 15935
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.57771913672781e-05,
+      "loss": 2.2141,
+      "step": 15936
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5776701971832187e-05,
+      "loss": 1.7332,
+      "step": 15937
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5776212555620395e-05,
+      "loss": 1.9346,
+      "step": 15938
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5775723118644483e-05,
+      "loss": 2.1577,
+      "step": 15939
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5775233660906213e-05,
+      "loss": 1.8226,
+      "step": 15940
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5774744182407346e-05,
+      "loss": 2.0981,
+      "step": 15941
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5774254683149636e-05,
+      "loss": 1.5066,
+      "step": 15942
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5773765163134848e-05,
+      "loss": 1.8994,
+      "step": 15943
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5773275622364744e-05,
+      "loss": 1.7289,
+      "step": 15944
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5772786060841073e-05,
+      "loss": 1.7055,
+      "step": 15945
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5772296478565603e-05,
+      "loss": 1.6435,
+      "step": 15946
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5771806875540092e-05,
+      "loss": 1.9354,
+      "step": 15947
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.57713172517663e-05,
+      "loss": 1.7606,
+      "step": 15948
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5770827607245992e-05,
+      "loss": 2.0811,
+      "step": 15949
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5770337941980917e-05,
+      "loss": 1.8509,
+      "step": 15950
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.576984825597285e-05,
+      "loss": 1.9072,
+      "step": 15951
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5769358549223532e-05,
+      "loss": 1.7456,
+      "step": 15952
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5768868821734743e-05,
+      "loss": 1.7539,
+      "step": 15953
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5768379073508232e-05,
+      "loss": 1.7736,
+      "step": 15954
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.576788930454576e-05,
+      "loss": 1.6368,
+      "step": 15955
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5767399514849097e-05,
+      "loss": 1.837,
+      "step": 15956
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.576690970441999e-05,
+      "loss": 1.5767,
+      "step": 15957
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5766419873260212e-05,
+      "loss": 1.945,
+      "step": 15958
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.576593002137151e-05,
+      "loss": 1.9666,
+      "step": 15959
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.576544014875566e-05,
+      "loss": 1.8533,
+      "step": 15960
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5764950255414415e-05,
+      "loss": 1.5978,
+      "step": 15961
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5764460341349533e-05,
+      "loss": 2.2918,
+      "step": 15962
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5763970406562785e-05,
+      "loss": 2.1447,
+      "step": 15963
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5763480451055925e-05,
+      "loss": 2.0664,
+      "step": 15964
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5762990474830716e-05,
+      "loss": 1.9136,
+      "step": 15965
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5762500477888915e-05,
+      "loss": 1.9385,
+      "step": 15966
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5762010460232292e-05,
+      "loss": 1.9321,
+      "step": 15967
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5761520421862604e-05,
+      "loss": 1.7851,
+      "step": 15968
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5761030362781612e-05,
+      "loss": 1.9108,
+      "step": 15969
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5760540282991077e-05,
+      "loss": 1.4539,
+      "step": 15970
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5760050182492762e-05,
+      "loss": 1.7051,
+      "step": 15971
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.575956006128843e-05,
+      "loss": 2.4711,
+      "step": 15972
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5759069919379842e-05,
+      "loss": 1.8969,
+      "step": 15973
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.575857975676876e-05,
+      "loss": 1.8871,
+      "step": 15974
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5758089573456942e-05,
+      "loss": 2.1599,
+      "step": 15975
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.575759936944616e-05,
+      "loss": 1.405,
+      "step": 15976
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5757109144738164e-05,
+      "loss": 1.8221,
+      "step": 15977
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.5756618899334725e-05,
+      "loss": 2.2512,
+      "step": 15978
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5756128633237603e-05,
+      "loss": 2.0115,
+      "step": 15979
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5755638346448557e-05,
+      "loss": 1.8465,
+      "step": 15980
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5755148038969356e-05,
+      "loss": 1.791,
+      "step": 15981
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5754657710801754e-05,
+      "loss": 1.5886,
+      "step": 15982
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.575416736194752e-05,
+      "loss": 1.9617,
+      "step": 15983
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.575367699240842e-05,
+      "loss": 1.7663,
+      "step": 15984
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5753186602186207e-05,
+      "loss": 1.6056,
+      "step": 15985
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.575269619128265e-05,
+      "loss": 2.0866,
+      "step": 15986
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.5752205759699512e-05,
+      "loss": 1.7489,
+      "step": 15987
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5751715307438555e-05,
+      "loss": 1.9385,
+      "step": 15988
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5751224834501537e-05,
+      "loss": 1.8683,
+      "step": 15989
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.575073434089023e-05,
+      "loss": 2.0092,
+      "step": 15990
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5750243826606393e-05,
+      "loss": 1.8998,
+      "step": 15991
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5749753291651788e-05,
+      "loss": 1.7859,
+      "step": 15992
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.574926273602818e-05,
+      "loss": 1.9541,
+      "step": 15993
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5748772159737333e-05,
+      "loss": 1.7569,
+      "step": 15994
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.574828156278101e-05,
+      "loss": 1.7507,
+      "step": 15995
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.5747790945160973e-05,
+      "loss": 1.8068,
+      "step": 15996
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5747300306878984e-05,
+      "loss": 1.6637,
+      "step": 15997
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5746809647936814e-05,
+      "loss": 1.7273,
+      "step": 15998
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5746318968336222e-05,
+      "loss": 2.0514,
+      "step": 15999
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.574582826807897e-05,
+      "loss": 1.962,
+      "step": 16000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.574533754716683e-05,
+      "loss": 2.0535,
+      "step": 16001
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.574484680560155e-05,
+      "loss": 2.0355,
+      "step": 16002
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5744356043384913e-05,
+      "loss": 1.8871,
+      "step": 16003
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5743865260518672e-05,
+      "loss": 1.9531,
+      "step": 16004
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5743374457004593e-05,
+      "loss": 2.0461,
+      "step": 16005
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5742883632844442e-05,
+      "loss": 1.9589,
+      "step": 16006
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.574239278803998e-05,
+      "loss": 1.7733,
+      "step": 16007
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5741901922592977e-05,
+      "loss": 1.8176,
+      "step": 16008
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.574141103650519e-05,
+      "loss": 2.0113,
+      "step": 16009
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5740920129778392e-05,
+      "loss": 1.9263,
+      "step": 16010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5740429202414347e-05,
+      "loss": 1.8864,
+      "step": 16011
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.573993825441481e-05,
+      "loss": 1.9283,
+      "step": 16012
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5739447285781556e-05,
+      "loss": 1.7471,
+      "step": 16013
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5738956296516347e-05,
+      "loss": 2.0406,
+      "step": 16014
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5738465286620946e-05,
+      "loss": 1.6709,
+      "step": 16015
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5737974256097118e-05,
+      "loss": 1.6899,
+      "step": 16016
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.573748320494663e-05,
+      "loss": 2.226,
+      "step": 16017
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.573699213317125e-05,
+      "loss": 1.7227,
+      "step": 16018
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5736501040772738e-05,
+      "loss": 1.9896,
+      "step": 16019
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.573600992775286e-05,
+      "loss": 1.8004,
+      "step": 16020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5735518794113386e-05,
+      "loss": 1.9005,
+      "step": 16021
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5735027639856076e-05,
+      "loss": 1.9499,
+      "step": 16022
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5734536464982695e-05,
+      "loss": 1.7719,
+      "step": 16023
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5734045269495016e-05,
+      "loss": 1.9803,
+      "step": 16024
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.57335540533948e-05,
+      "loss": 1.8933,
+      "step": 16025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5733062816683814e-05,
+      "loss": 1.8522,
+      "step": 16026
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.573257155936382e-05,
+      "loss": 1.7115,
+      "step": 16027
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.573208028143659e-05,
+      "loss": 1.7878,
+      "step": 16028
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5731588982903884e-05,
+      "loss": 1.6975,
+      "step": 16029
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5731097663767473e-05,
+      "loss": 1.8837,
+      "step": 16030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.573060632402912e-05,
+      "loss": 1.7608,
+      "step": 16031
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5730114963690596e-05,
+      "loss": 1.8372,
+      "step": 16032
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.572962358275366e-05,
+      "loss": 1.9106,
+      "step": 16033
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5729132181220083e-05,
+      "loss": 1.5905,
+      "step": 16034
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5728640759091632e-05,
+      "loss": 1.9319,
+      "step": 16035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.572814931637007e-05,
+      "loss": 1.7733,
+      "step": 16036
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5727657853057165e-05,
+      "loss": 1.9363,
+      "step": 16037
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5727166369154686e-05,
+      "loss": 1.8911,
+      "step": 16038
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.57266748646644e-05,
+      "loss": 1.8821,
+      "step": 16039
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5726183339588074e-05,
+      "loss": 2.2565,
+      "step": 16040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.572569179392747e-05,
+      "loss": 1.976,
+      "step": 16041
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.5725200227684357e-05,
+      "loss": 1.7808,
+      "step": 16042
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5724708640860504e-05,
+      "loss": 1.8566,
+      "step": 16043
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5724217033457677e-05,
+      "loss": 1.9083,
+      "step": 16044
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5723725405477645e-05,
+      "loss": 1.9781,
+      "step": 16045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5723233756922173e-05,
+      "loss": 1.7867,
+      "step": 16046
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.572274208779303e-05,
+      "loss": 1.7841,
+      "step": 16047
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5722250398091984e-05,
+      "loss": 2.1378,
+      "step": 16048
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.57217586878208e-05,
+      "loss": 1.5392,
+      "step": 16049
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5721266956981244e-05,
+      "loss": 1.8849,
+      "step": 16050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5720775205575088e-05,
+      "loss": 1.92,
+      "step": 16051
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.5720283433604097e-05,
+      "loss": 2.0157,
+      "step": 16052
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5719791641070042e-05,
+      "loss": 1.9229,
+      "step": 16053
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5719299827974687e-05,
+      "loss": 1.906,
+      "step": 16054
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.57188079943198e-05,
+      "loss": 2.1942,
+      "step": 16055
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5718316140107156e-05,
+      "loss": 1.7977,
+      "step": 16056
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.571782426533851e-05,
+      "loss": 1.9803,
+      "step": 16057
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5717332370015647e-05,
+      "loss": 1.7904,
+      "step": 16058
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.375,
+      "learning_rate": 1.571684045414032e-05,
+      "loss": 2.011,
+      "step": 16059
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5716348517714308e-05,
+      "loss": 1.7868,
+      "step": 16060
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5715856560739375e-05,
+      "loss": 1.5843,
+      "step": 16061
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5715364583217285e-05,
+      "loss": 2.017,
+      "step": 16062
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5714872585149812e-05,
+      "loss": 1.984,
+      "step": 16063
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5714380566538724e-05,
+      "loss": 1.7542,
+      "step": 16064
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5713888527385792e-05,
+      "loss": 1.5267,
+      "step": 16065
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.571339646769278e-05,
+      "loss": 1.6949,
+      "step": 16066
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5712904387461457e-05,
+      "loss": 1.8959,
+      "step": 16067
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5712412286693597e-05,
+      "loss": 1.9833,
+      "step": 16068
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5711920165390965e-05,
+      "loss": 2.1021,
+      "step": 16069
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.571142802355533e-05,
+      "loss": 1.7633,
+      "step": 16070
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5710935861188458e-05,
+      "loss": 1.9731,
+      "step": 16071
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.571044367829213e-05,
+      "loss": 1.9892,
+      "step": 16072
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5709951474868103e-05,
+      "loss": 2.224,
+      "step": 16073
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5709459250918153e-05,
+      "loss": 1.9045,
+      "step": 16074
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5708967006444045e-05,
+      "loss": 1.8539,
+      "step": 16075
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5708474741447552e-05,
+      "loss": 1.813,
+      "step": 16076
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5707982455930443e-05,
+      "loss": 1.7826,
+      "step": 16077
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5707490149894485e-05,
+      "loss": 2.0175,
+      "step": 16078
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5706997823341448e-05,
+      "loss": 1.6172,
+      "step": 16079
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5706505476273107e-05,
+      "loss": 1.7014,
+      "step": 16080
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5706013108691228e-05,
+      "loss": 1.857,
+      "step": 16081
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.570552072059758e-05,
+      "loss": 1.7338,
+      "step": 16082
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5705028311993937e-05,
+      "loss": 1.9167,
+      "step": 16083
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5704535882882063e-05,
+      "loss": 1.9136,
+      "step": 16084
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5704043433263737e-05,
+      "loss": 1.9037,
+      "step": 16085
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5703550963140717e-05,
+      "loss": 2.177,
+      "step": 16086
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5703058472514782e-05,
+      "loss": 1.4928,
+      "step": 16087
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5702565961387703e-05,
+      "loss": 1.9693,
+      "step": 16088
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.5702073429761247e-05,
+      "loss": 2.0934,
+      "step": 16089
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5701580877637184e-05,
+      "loss": 1.8151,
+      "step": 16090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5701088305017288e-05,
+      "loss": 2.5917,
+      "step": 16091
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5700595711903325e-05,
+      "loss": 1.8092,
+      "step": 16092
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5700103098297073e-05,
+      "loss": 2.0945,
+      "step": 16093
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5699610464200295e-05,
+      "loss": 1.8736,
+      "step": 16094
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.569911780961477e-05,
+      "loss": 1.7643,
+      "step": 16095
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5698625134542257e-05,
+      "loss": 2.3118,
+      "step": 16096
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.5698132438984538e-05,
+      "loss": 2.0621,
+      "step": 16097
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.569763972294338e-05,
+      "loss": 1.7867,
+      "step": 16098
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5697146986420555e-05,
+      "loss": 1.8705,
+      "step": 16099
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5696654229417833e-05,
+      "loss": 1.7354,
+      "step": 16100
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5696161451936988e-05,
+      "loss": 1.881,
+      "step": 16101
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.569566865397979e-05,
+      "loss": 1.9984,
+      "step": 16102
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5695175835548007e-05,
+      "loss": 2.02,
+      "step": 16103
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5694682996643416e-05,
+      "loss": 2.3261,
+      "step": 16104
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.569419013726779e-05,
+      "loss": 1.6995,
+      "step": 16105
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.569369725742289e-05,
+      "loss": 1.879,
+      "step": 16106
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.56932043571105e-05,
+      "loss": 2.25,
+      "step": 16107
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5692711436332383e-05,
+      "loss": 1.6725,
+      "step": 16108
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5692218495090315e-05,
+      "loss": 1.7607,
+      "step": 16109
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5691725533386072e-05,
+      "loss": 1.9348,
+      "step": 16110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.569123255122142e-05,
+      "loss": 1.8581,
+      "step": 16111
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.569073954859813e-05,
+      "loss": 1.7941,
+      "step": 16112
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5690246525517977e-05,
+      "loss": 1.9903,
+      "step": 16113
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5689753481982738e-05,
+      "loss": 1.982,
+      "step": 16114
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5689260417994178e-05,
+      "loss": 1.7522,
+      "step": 16115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.568876733355407e-05,
+      "loss": 1.8174,
+      "step": 16116
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.5688274228664197e-05,
+      "loss": 2.2673,
+      "step": 16117
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5687781103326317e-05,
+      "loss": 1.8125,
+      "step": 16118
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.568728795754221e-05,
+      "loss": 1.712,
+      "step": 16119
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5686794791313643e-05,
+      "loss": 1.8748,
+      "step": 16120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.56863016046424e-05,
+      "loss": 1.9094,
+      "step": 16121
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5685808397530248e-05,
+      "loss": 1.9679,
+      "step": 16122
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5685315169978955e-05,
+      "loss": 1.7103,
+      "step": 16123
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.56848219219903e-05,
+      "loss": 2.0168,
+      "step": 16124
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5684328653566056e-05,
+      "loss": 2.0698,
+      "step": 16125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5683835364707995e-05,
+      "loss": 1.7779,
+      "step": 16126
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.568334205541789e-05,
+      "loss": 1.9993,
+      "step": 16127
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.568284872569751e-05,
+      "loss": 2.0736,
+      "step": 16128
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.568235537554864e-05,
+      "loss": 1.805,
+      "step": 16129
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.568186200497304e-05,
+      "loss": 2.0227,
+      "step": 16130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5681368613972494e-05,
+      "loss": 1.7891,
+      "step": 16131
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5680875202548772e-05,
+      "loss": 1.7572,
+      "step": 16132
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.568038177070364e-05,
+      "loss": 1.8941,
+      "step": 16133
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5679888318438888e-05,
+      "loss": 1.851,
+      "step": 16134
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5679394845756277e-05,
+      "loss": 2.0429,
+      "step": 16135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5678901352657583e-05,
+      "loss": 1.7383,
+      "step": 16136
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.567840783914458e-05,
+      "loss": 1.9004,
+      "step": 16137
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5677914305219052e-05,
+      "loss": 1.8833,
+      "step": 16138
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5677420750882756e-05,
+      "loss": 2.004,
+      "step": 16139
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.567692717613748e-05,
+      "loss": 1.858,
+      "step": 16140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5676433580984992e-05,
+      "loss": 1.5043,
+      "step": 16141
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5675939965427066e-05,
+      "loss": 1.6914,
+      "step": 16142
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.567544632946548e-05,
+      "loss": 1.6021,
+      "step": 16143
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.567495267310201e-05,
+      "loss": 1.7628,
+      "step": 16144
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.567445899633842e-05,
+      "loss": 1.7968,
+      "step": 16145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.56739652991765e-05,
+      "loss": 1.7551,
+      "step": 16146
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5673471581618008e-05,
+      "loss": 1.7657,
+      "step": 16147
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5672977843664732e-05,
+      "loss": 1.9599,
+      "step": 16148
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5672484085318443e-05,
+      "loss": 2.1299,
+      "step": 16149
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5671990306580913e-05,
+      "loss": 1.7415,
+      "step": 16150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.567149650745392e-05,
+      "loss": 2.0322,
+      "step": 16151
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.567100268793924e-05,
+      "loss": 2.208,
+      "step": 16152
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5670508848038647e-05,
+      "loss": 2.0411,
+      "step": 16153
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.567001498775391e-05,
+      "loss": 1.8992,
+      "step": 16154
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5669521107086815e-05,
+      "loss": 1.6437,
+      "step": 16155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5669027206039134e-05,
+      "loss": 1.8722,
+      "step": 16156
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5668533284612636e-05,
+      "loss": 1.7962,
+      "step": 16157
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5668039342809104e-05,
+      "loss": 1.6552,
+      "step": 16158
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.566754538063031e-05,
+      "loss": 1.9355,
+      "step": 16159
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5667051398078035e-05,
+      "loss": 1.8207,
+      "step": 16160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5666557395154045e-05,
+      "loss": 2.0105,
+      "step": 16161
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5666063371860123e-05,
+      "loss": 1.7626,
+      "step": 16162
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5665569328198043e-05,
+      "loss": 1.9337,
+      "step": 16163
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.566507526416958e-05,
+      "loss": 2.0626,
+      "step": 16164
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5664581179776513e-05,
+      "loss": 1.9216,
+      "step": 16165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5664087075020616e-05,
+      "loss": 1.8274,
+      "step": 16166
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5663592949903667e-05,
+      "loss": 1.7228,
+      "step": 16167
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5663098804427438e-05,
+      "loss": 2.1669,
+      "step": 16168
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.566260463859371e-05,
+      "loss": 1.9381,
+      "step": 16169
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.566211045240426e-05,
+      "loss": 1.7741,
+      "step": 16170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5661616245860854e-05,
+      "loss": 1.8509,
+      "step": 16171
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5661122018965283e-05,
+      "loss": 1.6212,
+      "step": 16172
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5660627771719315e-05,
+      "loss": 1.9497,
+      "step": 16173
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.566013350412473e-05,
+      "loss": 1.6997,
+      "step": 16174
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.56596392161833e-05,
+      "loss": 1.7114,
+      "step": 16175
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.625,
+      "learning_rate": 1.565914490789681e-05,
+      "loss": 2.9048,
+      "step": 16176
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.565865057926703e-05,
+      "loss": 1.8966,
+      "step": 16177
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.565815623029574e-05,
+      "loss": 1.7088,
+      "step": 16178
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.565766186098472e-05,
+      "loss": 1.8663,
+      "step": 16179
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.565716747133574e-05,
+      "loss": 1.8797,
+      "step": 16180
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5656673061350578e-05,
+      "loss": 1.8334,
+      "step": 16181
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.565617863103102e-05,
+      "loss": 1.8554,
+      "step": 16182
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5655684180378833e-05,
+      "loss": 2.0065,
+      "step": 16183
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.56551897093958e-05,
+      "loss": 1.8916,
+      "step": 16184
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5654695218083697e-05,
+      "loss": 1.8614,
+      "step": 16185
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5654200706444306e-05,
+      "loss": 2.0599,
+      "step": 16186
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5653706174479394e-05,
+      "loss": 1.9913,
+      "step": 16187
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.565321162219075e-05,
+      "loss": 2.0865,
+      "step": 16188
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.565271704958015e-05,
+      "loss": 1.9268,
+      "step": 16189
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5652222456649363e-05,
+      "loss": 1.5963,
+      "step": 16190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5651727843400174e-05,
+      "loss": 1.9085,
+      "step": 16191
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.565123320983436e-05,
+      "loss": 1.591,
+      "step": 16192
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5650738555953703e-05,
+      "loss": 1.709,
+      "step": 16193
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5650243881759972e-05,
+      "loss": 1.9777,
+      "step": 16194
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5649749187254954e-05,
+      "loss": 1.8863,
+      "step": 16195
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5649254472440423e-05,
+      "loss": 1.8675,
+      "step": 16196
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5648759737318155e-05,
+      "loss": 1.9216,
+      "step": 16197
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5648264981889936e-05,
+      "loss": 1.8784,
+      "step": 16198
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.564777020615754e-05,
+      "loss": 1.81,
+      "step": 16199
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.564727541012274e-05,
+      "loss": 2.0305,
+      "step": 16200
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.564678059378733e-05,
+      "loss": 1.8223,
+      "step": 16201
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.564628575715307e-05,
+      "loss": 1.9589,
+      "step": 16202
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5645790900221752e-05,
+      "loss": 1.6666,
+      "step": 16203
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5645296022995152e-05,
+      "loss": 1.8378,
+      "step": 16204
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.564480112547505e-05,
+      "loss": 1.9079,
+      "step": 16205
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5644306207663217e-05,
+      "loss": 1.9833,
+      "step": 16206
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5643811269561443e-05,
+      "loss": 1.866,
+      "step": 16207
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.56433163111715e-05,
+      "loss": 2.0381,
+      "step": 16208
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.564282133249517e-05,
+      "loss": 1.8123,
+      "step": 16209
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.564232633353423e-05,
+      "loss": 1.9611,
+      "step": 16210
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5641831314290462e-05,
+      "loss": 1.9809,
+      "step": 16211
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5641336274765647e-05,
+      "loss": 2.0186,
+      "step": 16212
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.564084121496156e-05,
+      "loss": 1.8407,
+      "step": 16213
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.564034613487998e-05,
+      "loss": 2.1962,
+      "step": 16214
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5639851034522696e-05,
+      "loss": 1.8842,
+      "step": 16215
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.563935591389148e-05,
+      "loss": 1.9887,
+      "step": 16216
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.563886077298811e-05,
+      "loss": 1.8523,
+      "step": 16217
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5638365611814368e-05,
+      "loss": 2.042,
+      "step": 16218
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.563787043037204e-05,
+      "loss": 1.9435,
+      "step": 16219
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5637375228662896e-05,
+      "loss": 2.0622,
+      "step": 16220
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5636880006688726e-05,
+      "loss": 1.7468,
+      "step": 16221
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5636384764451302e-05,
+      "loss": 1.6589,
+      "step": 16222
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.563588950195241e-05,
+      "loss": 2.0379,
+      "step": 16223
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5635394219193827e-05,
+      "loss": 1.5144,
+      "step": 16224
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5634898916177334e-05,
+      "loss": 1.7632,
+      "step": 16225
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5634403592904712e-05,
+      "loss": 2.0077,
+      "step": 16226
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5633908249377743e-05,
+      "loss": 1.9725,
+      "step": 16227
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5633412885598206e-05,
+      "loss": 1.7634,
+      "step": 16228
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.563291750156788e-05,
+      "loss": 1.9289,
+      "step": 16229
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.563242209728855e-05,
+      "loss": 2.015,
+      "step": 16230
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5631926672761993e-05,
+      "loss": 1.9705,
+      "step": 16231
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5631431227989992e-05,
+      "loss": 1.887,
+      "step": 16232
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5630935762974328e-05,
+      "loss": 1.8241,
+      "step": 16233
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5630440277716782e-05,
+      "loss": 1.952,
+      "step": 16234
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5629944772219133e-05,
+      "loss": 1.912,
+      "step": 16235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5629449246483166e-05,
+      "loss": 1.9574,
+      "step": 16236
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.562895370051066e-05,
+      "loss": 1.7012,
+      "step": 16237
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5628458134303396e-05,
+      "loss": 1.6719,
+      "step": 16238
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5627962547863153e-05,
+      "loss": 1.8756,
+      "step": 16239
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.375,
+      "learning_rate": 1.562746694119172e-05,
+      "loss": 2.074,
+      "step": 16240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5626971314290872e-05,
+      "loss": 1.5644,
+      "step": 16241
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5626475667162396e-05,
+      "loss": 1.6849,
+      "step": 16242
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.562597999980807e-05,
+      "loss": 1.9557,
+      "step": 16243
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5625484312229674e-05,
+      "loss": 1.6694,
+      "step": 16244
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5624988604428993e-05,
+      "loss": 1.8254,
+      "step": 16245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.562449287640781e-05,
+      "loss": 1.7248,
+      "step": 16246
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5623997128167903e-05,
+      "loss": 1.7659,
+      "step": 16247
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5623501359711054e-05,
+      "loss": 1.844,
+      "step": 16248
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5623005571039052e-05,
+      "loss": 1.7055,
+      "step": 16249
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5622509762153676e-05,
+      "loss": 1.812,
+      "step": 16250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.5622013933056705e-05,
+      "loss": 1.968,
+      "step": 16251
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.562151808374992e-05,
+      "loss": 1.6863,
+      "step": 16252
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5621022214235112e-05,
+      "loss": 2.0363,
+      "step": 16253
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5620526324514055e-05,
+      "loss": 2.1937,
+      "step": 16254
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5620030414588535e-05,
+      "loss": 2.0804,
+      "step": 16255
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5619534484460336e-05,
+      "loss": 1.9553,
+      "step": 16256
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5619038534131243e-05,
+      "loss": 1.985,
+      "step": 16257
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.561854256360303e-05,
+      "loss": 1.9059,
+      "step": 16258
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5618046572877484e-05,
+      "loss": 1.6544,
+      "step": 16259
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.5617550561956397e-05,
+      "loss": 1.9432,
+      "step": 16260
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.561705453084154e-05,
+      "loss": 1.9381,
+      "step": 16261
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5616558479534697e-05,
+      "loss": 2.0052,
+      "step": 16262
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.561606240803766e-05,
+      "loss": 1.7744,
+      "step": 16263
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5615566316352203e-05,
+      "loss": 1.8045,
+      "step": 16264
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5615070204480113e-05,
+      "loss": 1.3459,
+      "step": 16265
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5614574072423177e-05,
+      "loss": 1.9136,
+      "step": 16266
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5614077920183174e-05,
+      "loss": 2.0824,
+      "step": 16267
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5613581747761886e-05,
+      "loss": 1.9637,
+      "step": 16268
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.56130855551611e-05,
+      "loss": 1.737,
+      "step": 16269
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.56125893423826e-05,
+      "loss": 1.9188,
+      "step": 16270
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.561209310942817e-05,
+      "loss": 1.8788,
+      "step": 16271
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.375,
+      "learning_rate": 1.561159685629959e-05,
+      "loss": 1.918,
+      "step": 16272
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.561110058299865e-05,
+      "loss": 1.741,
+      "step": 16273
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5610604289527124e-05,
+      "loss": 1.727,
+      "step": 16274
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5610107975886805e-05,
+      "loss": 2.1102,
+      "step": 16275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5609611642079476e-05,
+      "loss": 2.1003,
+      "step": 16276
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.560911528810692e-05,
+      "loss": 1.849,
+      "step": 16277
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.560861891397092e-05,
+      "loss": 1.9804,
+      "step": 16278
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.560812251967326e-05,
+      "loss": 1.737,
+      "step": 16279
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.560762610521573e-05,
+      "loss": 1.7957,
+      "step": 16280
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.560712967060011e-05,
+      "loss": 1.9084,
+      "step": 16281
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5606633215828184e-05,
+      "loss": 1.7616,
+      "step": 16282
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5606136740901736e-05,
+      "loss": 2.0292,
+      "step": 16283
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5605640245822552e-05,
+      "loss": 1.8327,
+      "step": 16284
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5605143730592417e-05,
+      "loss": 1.883,
+      "step": 16285
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5604647195213117e-05,
+      "loss": 1.9194,
+      "step": 16286
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5604150639686436e-05,
+      "loss": 1.6425,
+      "step": 16287
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5603654064014158e-05,
+      "loss": 1.8222,
+      "step": 16288
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.5603157468198067e-05,
+      "loss": 2.0386,
+      "step": 16289
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5602660852239953e-05,
+      "loss": 1.7733,
+      "step": 16290
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5602164216141598e-05,
+      "loss": 1.9975,
+      "step": 16291
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5601667559904783e-05,
+      "loss": 2.0878,
+      "step": 16292
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5601170883531306e-05,
+      "loss": 2.0432,
+      "step": 16293
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.560067418702294e-05,
+      "loss": 1.7403,
+      "step": 16294
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5600177470381472e-05,
+      "loss": 1.9311,
+      "step": 16295
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5599680733608694e-05,
+      "loss": 1.9369,
+      "step": 16296
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5599183976706388e-05,
+      "loss": 1.9196,
+      "step": 16297
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.559868719967634e-05,
+      "loss": 2.1068,
+      "step": 16298
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5598190402520335e-05,
+      "loss": 1.9767,
+      "step": 16299
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5597693585240158e-05,
+      "loss": 1.553,
+      "step": 16300
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5597196747837596e-05,
+      "loss": 2.0459,
+      "step": 16301
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5596699890314437e-05,
+      "loss": 1.8352,
+      "step": 16302
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5596203012672466e-05,
+      "loss": 1.6094,
+      "step": 16303
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.559570611491347e-05,
+      "loss": 1.7442,
+      "step": 16304
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.559520919703923e-05,
+      "loss": 2.0318,
+      "step": 16305
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5594712259051537e-05,
+      "loss": 1.8372,
+      "step": 16306
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.559421530095218e-05,
+      "loss": 1.94,
+      "step": 16307
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5593718322742936e-05,
+      "loss": 1.8297,
+      "step": 16308
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5593221324425602e-05,
+      "loss": 2.0135,
+      "step": 16309
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.559272430600196e-05,
+      "loss": 1.8723,
+      "step": 16310
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5592227267473794e-05,
+      "loss": 1.8967,
+      "step": 16311
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5591730208842896e-05,
+      "loss": 1.9209,
+      "step": 16312
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.559123313011105e-05,
+      "loss": 1.7839,
+      "step": 16313
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5590736031280044e-05,
+      "loss": 1.6825,
+      "step": 16314
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5590238912351662e-05,
+      "loss": 1.9452,
+      "step": 16315
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5589741773327694e-05,
+      "loss": 1.9142,
+      "step": 16316
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.558924461420993e-05,
+      "loss": 1.7372,
+      "step": 16317
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5588747435000144e-05,
+      "loss": 2.1313,
+      "step": 16318
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5588250235700143e-05,
+      "loss": 1.782,
+      "step": 16319
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5587753016311698e-05,
+      "loss": 1.6833,
+      "step": 16320
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5587255776836605e-05,
+      "loss": 2.0625,
+      "step": 16321
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5586758517276644e-05,
+      "loss": 1.7622,
+      "step": 16322
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5586261237633613e-05,
+      "loss": 2.0122,
+      "step": 16323
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5585763937909294e-05,
+      "loss": 1.822,
+      "step": 16324
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.558526661810547e-05,
+      "loss": 1.941,
+      "step": 16325
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5584769278223935e-05,
+      "loss": 1.6891,
+      "step": 16326
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.558427191826648e-05,
+      "loss": 1.9065,
+      "step": 16327
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5583774538234883e-05,
+      "loss": 2.1667,
+      "step": 16328
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5583277138130935e-05,
+      "loss": 2.0186,
+      "step": 16329
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5582779717956434e-05,
+      "loss": 2.007,
+      "step": 16330
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5582282277713154e-05,
+      "loss": 1.7161,
+      "step": 16331
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.558178481740289e-05,
+      "loss": 1.9336,
+      "step": 16332
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.558128733702743e-05,
+      "loss": 1.6538,
+      "step": 16333
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5580789836588565e-05,
+      "loss": 1.3497,
+      "step": 16334
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5580292316088076e-05,
+      "loss": 1.7067,
+      "step": 16335
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.557979477552776e-05,
+      "loss": 1.7737,
+      "step": 16336
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.55792972149094e-05,
+      "loss": 1.8499,
+      "step": 16337
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5578799634234782e-05,
+      "loss": 1.8703,
+      "step": 16338
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5578302033505705e-05,
+      "loss": 2.0287,
+      "step": 16339
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.557780441272395e-05,
+      "loss": 1.8734,
+      "step": 16340
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5577306771891303e-05,
+      "loss": 1.7659,
+      "step": 16341
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5576809111009557e-05,
+      "loss": 1.8468,
+      "step": 16342
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5576311430080507e-05,
+      "loss": 2.2579,
+      "step": 16343
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.557581372910593e-05,
+      "loss": 1.9078,
+      "step": 16344
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5575316008087623e-05,
+      "loss": 1.7399,
+      "step": 16345
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5574818267027375e-05,
+      "loss": 1.5692,
+      "step": 16346
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5574320505926978e-05,
+      "loss": 2.0265,
+      "step": 16347
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.557382272478821e-05,
+      "loss": 1.8307,
+      "step": 16348
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5573324923612872e-05,
+      "loss": 1.7229,
+      "step": 16349
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5572827102402744e-05,
+      "loss": 1.949,
+      "step": 16350
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5572329261159627e-05,
+      "loss": 1.7064,
+      "step": 16351
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5571831399885298e-05,
+      "loss": 1.9347,
+      "step": 16352
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5571333518581554e-05,
+      "loss": 1.8707,
+      "step": 16353
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5570835617250187e-05,
+      "loss": 2.0804,
+      "step": 16354
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5570337695892983e-05,
+      "loss": 1.9806,
+      "step": 16355
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.556983975451173e-05,
+      "loss": 2.1984,
+      "step": 16356
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.556934179310822e-05,
+      "loss": 1.5147,
+      "step": 16357
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5568843811684246e-05,
+      "loss": 2.0962,
+      "step": 16358
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.556834581024159e-05,
+      "loss": 1.8789,
+      "step": 16359
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5567847788782057e-05,
+      "loss": 1.8628,
+      "step": 16360
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.556734974730742e-05,
+      "loss": 2.2168,
+      "step": 16361
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5566851685819478e-05,
+      "loss": 1.8654,
+      "step": 16362
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5566353604320022e-05,
+      "loss": 2.0438,
+      "step": 16363
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5565855502810846e-05,
+      "loss": 2.0031,
+      "step": 16364
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.5565357381293732e-05,
+      "loss": 2.0015,
+      "step": 16365
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5564859239770472e-05,
+      "loss": 2.1302,
+      "step": 16366
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5564361078242862e-05,
+      "loss": 1.8345,
+      "step": 16367
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5563862896712688e-05,
+      "loss": 1.9094,
+      "step": 16368
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.556336469518174e-05,
+      "loss": 1.8568,
+      "step": 16369
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5562866473651815e-05,
+      "loss": 1.8701,
+      "step": 16370
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5562368232124705e-05,
+      "loss": 1.9081,
+      "step": 16371
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5561869970602192e-05,
+      "loss": 1.661,
+      "step": 16372
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.556137168908607e-05,
+      "loss": 2.1064,
+      "step": 16373
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5560873387578138e-05,
+      "loss": 1.9085,
+      "step": 16374
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5560375066080177e-05,
+      "loss": 1.7545,
+      "step": 16375
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.555987672459398e-05,
+      "loss": 1.771,
+      "step": 16376
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5559378363121346e-05,
+      "loss": 1.8625,
+      "step": 16377
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.555887998166406e-05,
+      "loss": 2.1038,
+      "step": 16378
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5558381580223913e-05,
+      "loss": 1.9254,
+      "step": 16379
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.55578831588027e-05,
+      "loss": 1.8037,
+      "step": 16380
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5557384717402216e-05,
+      "loss": 2.1974,
+      "step": 16381
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.375,
+      "learning_rate": 1.555688625602424e-05,
+      "loss": 1.8107,
+      "step": 16382
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.555638777467058e-05,
+      "loss": 2.133,
+      "step": 16383
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5555889273343015e-05,
+      "loss": 1.9995,
+      "step": 16384
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5555390752043345e-05,
+      "loss": 1.9684,
+      "step": 16385
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5554892210773355e-05,
+      "loss": 1.912,
+      "step": 16386
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5554393649534848e-05,
+      "loss": 1.8132,
+      "step": 16387
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5553895068329602e-05,
+      "loss": 1.7466,
+      "step": 16388
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5553396467159418e-05,
+      "loss": 1.8483,
+      "step": 16389
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.555289784602609e-05,
+      "loss": 1.672,
+      "step": 16390
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5552399204931406e-05,
+      "loss": 1.6979,
+      "step": 16391
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.555190054387716e-05,
+      "loss": 1.6683,
+      "step": 16392
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5551401862865146e-05,
+      "loss": 2.273,
+      "step": 16393
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5550903161897153e-05,
+      "loss": 1.8135,
+      "step": 16394
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5550404440974975e-05,
+      "loss": 1.565,
+      "step": 16395
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.554990570010041e-05,
+      "loss": 2.0765,
+      "step": 16396
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5549406939275242e-05,
+      "loss": 1.7011,
+      "step": 16397
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5548908158501272e-05,
+      "loss": 1.8814,
+      "step": 16398
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5548409357780286e-05,
+      "loss": 2.0123,
+      "step": 16399
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5547910537114083e-05,
+      "loss": 2.0727,
+      "step": 16400
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5547411696504452e-05,
+      "loss": 1.7603,
+      "step": 16401
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.554691283595319e-05,
+      "loss": 1.7476,
+      "step": 16402
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.554641395546209e-05,
+      "loss": 1.7898,
+      "step": 16403
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5545915055032937e-05,
+      "loss": 1.9766,
+      "step": 16404
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5545416134667538e-05,
+      "loss": 2.0701,
+      "step": 16405
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.5544917194367672e-05,
+      "loss": 2.7023,
+      "step": 16406
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5544418234135145e-05,
+      "loss": 1.7369,
+      "step": 16407
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5543919253971747e-05,
+      "loss": 1.831,
+      "step": 16408
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5543420253879267e-05,
+      "loss": 1.8588,
+      "step": 16409
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5542921233859502e-05,
+      "loss": 1.7043,
+      "step": 16410
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.554242219391425e-05,
+      "loss": 1.6711,
+      "step": 16411
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5541923134045296e-05,
+      "loss": 1.9128,
+      "step": 16412
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.554142405425444e-05,
+      "loss": 1.9039,
+      "step": 16413
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.554092495454348e-05,
+      "loss": 1.8286,
+      "step": 16414
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5540425834914202e-05,
+      "loss": 1.8847,
+      "step": 16415
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.55399266953684e-05,
+      "loss": 1.7738,
+      "step": 16416
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5539427535907872e-05,
+      "loss": 1.8732,
+      "step": 16417
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.553892835653442e-05,
+      "loss": 1.5084,
+      "step": 16418
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5538429157249823e-05,
+      "loss": 1.7197,
+      "step": 16419
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5537929938055882e-05,
+      "loss": 2.0579,
+      "step": 16420
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.55374306989544e-05,
+      "loss": 1.8925,
+      "step": 16421
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.553693143994716e-05,
+      "loss": 1.7684,
+      "step": 16422
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5536432161035956e-05,
+      "loss": 1.8649,
+      "step": 16423
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5535932862222595e-05,
+      "loss": 2.003,
+      "step": 16424
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5535433543508862e-05,
+      "loss": 1.9337,
+      "step": 16425
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5534934204896554e-05,
+      "loss": 1.8,
+      "step": 16426
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.553443484638747e-05,
+      "loss": 1.9202,
+      "step": 16427
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5533935467983396e-05,
+      "loss": 1.7299,
+      "step": 16428
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5533436069686134e-05,
+      "loss": 1.9612,
+      "step": 16429
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.553293665149748e-05,
+      "loss": 2.1002,
+      "step": 16430
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5532437213419226e-05,
+      "loss": 1.6573,
+      "step": 16431
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.553193775545317e-05,
+      "loss": 2.273,
+      "step": 16432
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5531438277601102e-05,
+      "loss": 2.0947,
+      "step": 16433
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5530938779864828e-05,
+      "loss": 1.826,
+      "step": 16434
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5530439262246133e-05,
+      "loss": 1.9871,
+      "step": 16435
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5529939724746816e-05,
+      "loss": 1.9142,
+      "step": 16436
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5529440167368676e-05,
+      "loss": 1.9102,
+      "step": 16437
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5528940590113505e-05,
+      "loss": 1.6134,
+      "step": 16438
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5528440992983102e-05,
+      "loss": 2.0706,
+      "step": 16439
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.552794137597926e-05,
+      "loss": 1.8817,
+      "step": 16440
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5527441739103774e-05,
+      "loss": 2.0382,
+      "step": 16441
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5526942082358444e-05,
+      "loss": 1.8356,
+      "step": 16442
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5526442405745064e-05,
+      "loss": 1.8247,
+      "step": 16443
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5525942709265432e-05,
+      "loss": 1.7111,
+      "step": 16444
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.552544299292134e-05,
+      "loss": 1.8779,
+      "step": 16445
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5524943256714587e-05,
+      "loss": 1.8637,
+      "step": 16446
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5524443500646973e-05,
+      "loss": 2.1567,
+      "step": 16447
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.552394372472029e-05,
+      "loss": 1.9566,
+      "step": 16448
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5523443928936336e-05,
+      "loss": 1.9338,
+      "step": 16449
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.55229441132969e-05,
+      "loss": 1.8192,
+      "step": 16450
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5522444277803797e-05,
+      "loss": 2.0002,
+      "step": 16451
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5521944422458806e-05,
+      "loss": 2.0017,
+      "step": 16452
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.5521444547263732e-05,
+      "loss": 1.7252,
+      "step": 16453
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5520944652220373e-05,
+      "loss": 1.936,
+      "step": 16454
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5520444737330524e-05,
+      "loss": 2.0841,
+      "step": 16455
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.551994480259598e-05,
+      "loss": 2.1031,
+      "step": 16456
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5519444848018537e-05,
+      "loss": 2.1969,
+      "step": 16457
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.55189448736e-05,
+      "loss": 1.8911,
+      "step": 16458
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.551844487934216e-05,
+      "loss": 1.8508,
+      "step": 16459
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.551794486524681e-05,
+      "loss": 1.6657,
+      "step": 16460
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.551744483131576e-05,
+      "loss": 1.8696,
+      "step": 16461
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5516944777550798e-05,
+      "loss": 2.1422,
+      "step": 16462
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5516444703953727e-05,
+      "loss": 2.1108,
+      "step": 16463
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5515944610526338e-05,
+      "loss": 2.2271,
+      "step": 16464
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5515444497270435e-05,
+      "loss": 1.7419,
+      "step": 16465
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5514944364187812e-05,
+      "loss": 1.7284,
+      "step": 16466
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5514444211280272e-05,
+      "loss": 1.5659,
+      "step": 16467
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5513944038549605e-05,
+      "loss": 1.5421,
+      "step": 16468
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5513443845997613e-05,
+      "loss": 1.7257,
+      "step": 16469
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.55129436336261e-05,
+      "loss": 2.0572,
+      "step": 16470
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.551244340143685e-05,
+      "loss": 1.5884,
+      "step": 16471
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5511943149431673e-05,
+      "loss": 1.7326,
+      "step": 16472
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5511442877612365e-05,
+      "loss": 1.9741,
+      "step": 16473
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5510942585980722e-05,
+      "loss": 2.1424,
+      "step": 16474
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5510442274538548e-05,
+      "loss": 1.9839,
+      "step": 16475
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5509941943287634e-05,
+      "loss": 1.6815,
+      "step": 16476
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.550944159222978e-05,
+      "loss": 2.0418,
+      "step": 16477
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.550894122136679e-05,
+      "loss": 1.6754,
+      "step": 16478
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5508440830700456e-05,
+      "loss": 2.0121,
+      "step": 16479
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.550794042023258e-05,
+      "loss": 1.9731,
+      "step": 16480
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.5507439989964966e-05,
+      "loss": 1.9169,
+      "step": 16481
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5506939539899403e-05,
+      "loss": 1.8707,
+      "step": 16482
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5506439070037696e-05,
+      "loss": 1.8466,
+      "step": 16483
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.550593858038164e-05,
+      "loss": 1.7943,
+      "step": 16484
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5505438070933042e-05,
+      "loss": 1.7108,
+      "step": 16485
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.550493754169369e-05,
+      "loss": 1.7575,
+      "step": 16486
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5504436992665394e-05,
+      "loss": 1.9071,
+      "step": 16487
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5503936423849948e-05,
+      "loss": 2.0985,
+      "step": 16488
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.550343583524915e-05,
+      "loss": 1.6664,
+      "step": 16489
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5502935226864804e-05,
+      "loss": 1.8794,
+      "step": 16490
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5502434598698705e-05,
+      "loss": 1.7142,
+      "step": 16491
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5501933950752655e-05,
+      "loss": 1.9294,
+      "step": 16492
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5501433283028457e-05,
+      "loss": 2.2735,
+      "step": 16493
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5500932595527903e-05,
+      "loss": 2.0964,
+      "step": 16494
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5500431888252806e-05,
+      "loss": 2.0862,
+      "step": 16495
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5499931161204948e-05,
+      "loss": 1.9827,
+      "step": 16496
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5499430414386138e-05,
+      "loss": 1.8395,
+      "step": 16497
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.549892964779818e-05,
+      "loss": 1.637,
+      "step": 16498
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5498428861442872e-05,
+      "loss": 1.7641,
+      "step": 16499
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5497928055322008e-05,
+      "loss": 2.0063,
+      "step": 16500
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5497427229437396e-05,
+      "loss": 1.933,
+      "step": 16501
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5496926383790833e-05,
+      "loss": 1.9825,
+      "step": 16502
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.549642551838412e-05,
+      "loss": 1.8872,
+      "step": 16503
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5495924633219054e-05,
+      "loss": 2.1794,
+      "step": 16504
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5495423728297445e-05,
+      "loss": 1.9763,
+      "step": 16505
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.549492280362108e-05,
+      "loss": 2.0366,
+      "step": 16506
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.549442185919177e-05,
+      "loss": 2.0047,
+      "step": 16507
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.5493920895011315e-05,
+      "loss": 1.8544,
+      "step": 16508
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.549341991108151e-05,
+      "loss": 2.0406,
+      "step": 16509
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.549291890740416e-05,
+      "loss": 1.7057,
+      "step": 16510
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5492417883981067e-05,
+      "loss": 1.6108,
+      "step": 16511
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.549191684081403e-05,
+      "loss": 1.9495,
+      "step": 16512
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5491415777904854e-05,
+      "loss": 1.9185,
+      "step": 16513
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5490914695255332e-05,
+      "loss": 1.6359,
+      "step": 16514
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5490413592867276e-05,
+      "loss": 2.2727,
+      "step": 16515
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5489912470742475e-05,
+      "loss": 1.9357,
+      "step": 16516
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.548941132888274e-05,
+      "loss": 1.8349,
+      "step": 16517
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.548891016728987e-05,
+      "loss": 1.8092,
+      "step": 16518
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5488408985965664e-05,
+      "loss": 1.9196,
+      "step": 16519
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5487907784911927e-05,
+      "loss": 1.9118,
+      "step": 16520
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.548740656413046e-05,
+      "loss": 1.8264,
+      "step": 16521
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5486905323623065e-05,
+      "loss": 2.1426,
+      "step": 16522
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.548640406339154e-05,
+      "loss": 1.9571,
+      "step": 16523
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5485902783437692e-05,
+      "loss": 1.931,
+      "step": 16524
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5485401483763324e-05,
+      "loss": 1.9849,
+      "step": 16525
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5484900164370228e-05,
+      "loss": 1.0166,
+      "step": 16526
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5484398825260218e-05,
+      "loss": 1.801,
+      "step": 16527
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.548389746643509e-05,
+      "loss": 1.6508,
+      "step": 16528
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5483396087896643e-05,
+      "loss": 2.0005,
+      "step": 16529
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.548289468964669e-05,
+      "loss": 1.6896,
+      "step": 16530
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.5482393271687028e-05,
+      "loss": 2.1006,
+      "step": 16531
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5481891834019452e-05,
+      "loss": 1.8076,
+      "step": 16532
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5481390376645777e-05,
+      "loss": 1.8546,
+      "step": 16533
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5480888899567797e-05,
+      "loss": 1.4349,
+      "step": 16534
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.548038740278732e-05,
+      "loss": 2.0477,
+      "step": 16535
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5479885886306144e-05,
+      "loss": 1.6666,
+      "step": 16536
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5479384350126076e-05,
+      "loss": 2.195,
+      "step": 16537
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.5478882794248916e-05,
+      "loss": 2.2219,
+      "step": 16538
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.547838121867647e-05,
+      "loss": 1.6437,
+      "step": 16539
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.5477879623410536e-05,
+      "loss": 2.0163,
+      "step": 16540
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.547737800845292e-05,
+      "loss": 1.8343,
+      "step": 16541
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.547687637380543e-05,
+      "loss": 1.9511,
+      "step": 16542
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5476374719469862e-05,
+      "loss": 1.9202,
+      "step": 16543
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.547587304544802e-05,
+      "loss": 1.6924,
+      "step": 16544
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5475371351741712e-05,
+      "loss": 2.0305,
+      "step": 16545
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5474869638352742e-05,
+      "loss": 2.2863,
+      "step": 16546
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5474367905282903e-05,
+      "loss": 2.0373,
+      "step": 16547
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.547386615253401e-05,
+      "loss": 1.496,
+      "step": 16548
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5473364380107863e-05,
+      "loss": 1.7171,
+      "step": 16549
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5472862588006265e-05,
+      "loss": 1.7775,
+      "step": 16550
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.547236077623102e-05,
+      "loss": 1.7687,
+      "step": 16551
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5471858944783933e-05,
+      "loss": 1.7222,
+      "step": 16552
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5471357093666805e-05,
+      "loss": 1.6874,
+      "step": 16553
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5470855222881444e-05,
+      "loss": 1.8717,
+      "step": 16554
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5470353332429652e-05,
+      "loss": 1.8233,
+      "step": 16555
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5469851422313235e-05,
+      "loss": 2.0938,
+      "step": 16556
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5469349492533994e-05,
+      "loss": 1.5747,
+      "step": 16557
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.5468847543093733e-05,
+      "loss": 2.3511,
+      "step": 16558
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5468345573994264e-05,
+      "loss": 2.0701,
+      "step": 16559
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.546784358523738e-05,
+      "loss": 1.78,
+      "step": 16560
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5467341576824893e-05,
+      "loss": 1.846,
+      "step": 16561
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.546683954875861e-05,
+      "loss": 2.0715,
+      "step": 16562
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5466337501040326e-05,
+      "loss": 1.7527,
+      "step": 16563
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5465835433671856e-05,
+      "loss": 2.0879,
+      "step": 16564
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5465333346655e-05,
+      "loss": 1.9426,
+      "step": 16565
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.546483123999156e-05,
+      "loss": 1.8937,
+      "step": 16566
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5464329113683344e-05,
+      "loss": 2.0914,
+      "step": 16567
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5463826967732158e-05,
+      "loss": 1.8722,
+      "step": 16568
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.546332480213981e-05,
+      "loss": 1.5897,
+      "step": 16569
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5462822616908096e-05,
+      "loss": 1.8941,
+      "step": 16570
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.546232041203883e-05,
+      "loss": 1.8767,
+      "step": 16571
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.5461818187533814e-05,
+      "loss": 1.689,
+      "step": 16572
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.546131594339485e-05,
+      "loss": 1.8325,
+      "step": 16573
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5460813679623748e-05,
+      "loss": 2.4005,
+      "step": 16574
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5460311396222313e-05,
+      "loss": 1.6855,
+      "step": 16575
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.5459809093192352e-05,
+      "loss": 1.9053,
+      "step": 16576
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5459306770535666e-05,
+      "loss": 1.87,
+      "step": 16577
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.545880442825406e-05,
+      "loss": 1.9744,
+      "step": 16578
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.545830206634935e-05,
+      "loss": 1.9141,
+      "step": 16579
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.5457799684823333e-05,
+      "loss": 1.9921,
+      "step": 16580
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.5457297283677816e-05,
+      "loss": 1.585,
+      "step": 16581
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5456794862914603e-05,
+      "loss": 1.957,
+      "step": 16582
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5456292422535506e-05,
+      "loss": 2.1536,
+      "step": 16583
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5455789962542327e-05,
+      "loss": 1.9372,
+      "step": 16584
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5455287482936877e-05,
+      "loss": 1.7972,
+      "step": 16585
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5454784983720954e-05,
+      "loss": 2.0795,
+      "step": 16586
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.545428246489637e-05,
+      "loss": 1.8047,
+      "step": 16587
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.545377992646493e-05,
+      "loss": 1.7199,
+      "step": 16588
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5453277368428443e-05,
+      "loss": 1.8975,
+      "step": 16589
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.545277479078871e-05,
+      "loss": 1.5978,
+      "step": 16590
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5452272193547545e-05,
+      "loss": 1.9766,
+      "step": 16591
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.545176957670675e-05,
+      "loss": 1.2389,
+      "step": 16592
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5451266940268133e-05,
+      "loss": 1.5976,
+      "step": 16593
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.54507642842335e-05,
+      "loss": 2.1944,
+      "step": 16594
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5450261608604655e-05,
+      "loss": 1.7877,
+      "step": 16595
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5449758913383414e-05,
+      "loss": 2.0646,
+      "step": 16596
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5449256198571575e-05,
+      "loss": 1.9543,
+      "step": 16597
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.544875346417095e-05,
+      "loss": 1.5068,
+      "step": 16598
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5448250710183345e-05,
+      "loss": 1.765,
+      "step": 16599
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5447747936610567e-05,
+      "loss": 2.1778,
+      "step": 16600
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5447245143454425e-05,
+      "loss": 1.9075,
+      "step": 16601
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5446742330716724e-05,
+      "loss": 1.8917,
+      "step": 16602
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.544623949839927e-05,
+      "loss": 1.7502,
+      "step": 16603
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5445736646503876e-05,
+      "loss": 1.72,
+      "step": 16604
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5445233775032344e-05,
+      "loss": 1.698,
+      "step": 16605
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5444730883986484e-05,
+      "loss": 1.8383,
+      "step": 16606
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.544422797336811e-05,
+      "loss": 1.9655,
+      "step": 16607
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.544372504317902e-05,
+      "loss": 2.0141,
+      "step": 16608
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5443222093421027e-05,
+      "loss": 1.8794,
+      "step": 16609
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.5442719124095937e-05,
+      "loss": 2.1984,
+      "step": 16610
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5442216135205554e-05,
+      "loss": 1.9565,
+      "step": 16611
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5441713126751698e-05,
+      "loss": 2.1567,
+      "step": 16612
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.544121009873617e-05,
+      "loss": 1.9907,
+      "step": 16613
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.544070705116077e-05,
+      "loss": 1.8212,
+      "step": 16614
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5440203984027323e-05,
+      "loss": 2.3627,
+      "step": 16615
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.543970089733763e-05,
+      "loss": 1.7382,
+      "step": 16616
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5439197791093493e-05,
+      "loss": 2.8007,
+      "step": 16617
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.543869466529673e-05,
+      "loss": 1.6254,
+      "step": 16618
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5438191519949145e-05,
+      "loss": 2.117,
+      "step": 16619
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.543768835505255e-05,
+      "loss": 2.3931,
+      "step": 16620
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5437185170608747e-05,
+      "loss": 1.7221,
+      "step": 16621
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.543668196661955e-05,
+      "loss": 1.8293,
+      "step": 16622
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.543617874308677e-05,
+      "loss": 1.8641,
+      "step": 16623
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5435675500012212e-05,
+      "loss": 1.7079,
+      "step": 16624
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5435172237397688e-05,
+      "loss": 1.4699,
+      "step": 16625
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5434668955245e-05,
+      "loss": 1.6333,
+      "step": 16626
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5434165653555968e-05,
+      "loss": 2.037,
+      "step": 16627
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.543366233233239e-05,
+      "loss": 1.8693,
+      "step": 16628
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5433158991576086e-05,
+      "loss": 1.9276,
+      "step": 16629
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5432655631288857e-05,
+      "loss": 1.9765,
+      "step": 16630
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5432152251472515e-05,
+      "loss": 1.7636,
+      "step": 16631
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5431648852128875e-05,
+      "loss": 1.9021,
+      "step": 16632
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5431145433259736e-05,
+      "loss": 1.7815,
+      "step": 16633
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.543064199486692e-05,
+      "loss": 2.0778,
+      "step": 16634
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5430138536952224e-05,
+      "loss": 1.8187,
+      "step": 16635
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.542963505951747e-05,
+      "loss": 1.7411,
+      "step": 16636
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5429131562564456e-05,
+      "loss": 2.0688,
+      "step": 16637
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5428628046095004e-05,
+      "loss": 1.745,
+      "step": 16638
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5428124510110916e-05,
+      "loss": 1.9036,
+      "step": 16639
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5427620954614e-05,
+      "loss": 1.883,
+      "step": 16640
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5427117379606075e-05,
+      "loss": 1.8656,
+      "step": 16641
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5426613785088945e-05,
+      "loss": 1.622,
+      "step": 16642
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5426110171064425e-05,
+      "loss": 1.7117,
+      "step": 16643
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5425606537534318e-05,
+      "loss": 2.0289,
+      "step": 16644
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5425102884500434e-05,
+      "loss": 1.7596,
+      "step": 16645
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5424599211964596e-05,
+      "loss": 1.885,
+      "step": 16646
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5424095519928606e-05,
+      "loss": 1.8342,
+      "step": 16647
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5423591808394273e-05,
+      "loss": 1.9335,
+      "step": 16648
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5423088077363414e-05,
+      "loss": 1.6539,
+      "step": 16649
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.542258432683783e-05,
+      "loss": 1.7269,
+      "step": 16650
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5422080556819342e-05,
+      "loss": 1.9375,
+      "step": 16651
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5421576767309755e-05,
+      "loss": 1.8405,
+      "step": 16652
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5421072958310885e-05,
+      "loss": 1.6831,
+      "step": 16653
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5420569129824534e-05,
+      "loss": 1.8666,
+      "step": 16654
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.542006528185252e-05,
+      "loss": 1.7802,
+      "step": 16655
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5419561414396657e-05,
+      "loss": 1.9635,
+      "step": 16656
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5419057527458748e-05,
+      "loss": 1.8012,
+      "step": 16657
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.541855362104061e-05,
+      "loss": 1.8311,
+      "step": 16658
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5418049695144054e-05,
+      "loss": 1.7601,
+      "step": 16659
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5417545749770894e-05,
+      "loss": 2.0454,
+      "step": 16660
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5417041784922932e-05,
+      "loss": 1.8275,
+      "step": 16661
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.541653780060199e-05,
+      "loss": 1.9141,
+      "step": 16662
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5416033796809877e-05,
+      "loss": 1.6185,
+      "step": 16663
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5415529773548398e-05,
+      "loss": 1.8967,
+      "step": 16664
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5415025730819372e-05,
+      "loss": 1.9887,
+      "step": 16665
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5414521668624614e-05,
+      "loss": 1.5678,
+      "step": 16666
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5414017586965926e-05,
+      "loss": 1.8447,
+      "step": 16667
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5413513485845125e-05,
+      "loss": 1.9003,
+      "step": 16668
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5413009365264023e-05,
+      "loss": 1.6942,
+      "step": 16669
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.541250522522444e-05,
+      "loss": 1.9583,
+      "step": 16670
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.541200106572817e-05,
+      "loss": 2.0161,
+      "step": 16671
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5411496886777043e-05,
+      "loss": 1.9097,
+      "step": 16672
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5410992688372865e-05,
+      "loss": 1.7946,
+      "step": 16673
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5410488470517442e-05,
+      "loss": 1.8066,
+      "step": 16674
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5409984233212595e-05,
+      "loss": 2.02,
+      "step": 16675
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5409479976460135e-05,
+      "loss": 1.8758,
+      "step": 16676
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5408975700261876e-05,
+      "loss": 1.5529,
+      "step": 16677
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5408471404619623e-05,
+      "loss": 1.683,
+      "step": 16678
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.54079670895352e-05,
+      "loss": 1.7219,
+      "step": 16679
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.540746275501041e-05,
+      "loss": 1.8764,
+      "step": 16680
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.5406958401047073e-05,
+      "loss": 1.6011,
+      "step": 16681
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5406454027647e-05,
+      "loss": 1.7462,
+      "step": 16682
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5405949634812002e-05,
+      "loss": 1.8422,
+      "step": 16683
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5405445222543893e-05,
+      "loss": 1.9915,
+      "step": 16684
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5404940790844486e-05,
+      "loss": 2.2186,
+      "step": 16685
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5404436339715597e-05,
+      "loss": 1.6453,
+      "step": 16686
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.540393186915904e-05,
+      "loss": 1.8205,
+      "step": 16687
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.540342737917662e-05,
+      "loss": 1.9445,
+      "step": 16688
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5402922869770156e-05,
+      "loss": 1.7546,
+      "step": 16689
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5402418340941467e-05,
+      "loss": 1.9891,
+      "step": 16690
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5401913792692356e-05,
+      "loss": 1.8715,
+      "step": 16691
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.540140922502465e-05,
+      "loss": 1.9645,
+      "step": 16692
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.540090463794015e-05,
+      "loss": 1.9121,
+      "step": 16693
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5400400031440676e-05,
+      "loss": 1.849,
+      "step": 16694
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5399895405528043e-05,
+      "loss": 1.9408,
+      "step": 16695
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5399390760204064e-05,
+      "loss": 2.2369,
+      "step": 16696
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5398886095470548e-05,
+      "loss": 1.536,
+      "step": 16697
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5398381411329312e-05,
+      "loss": 1.8428,
+      "step": 16698
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5397876707782176e-05,
+      "loss": 2.0527,
+      "step": 16699
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.539737198483095e-05,
+      "loss": 1.7777,
+      "step": 16700
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5396867242477447e-05,
+      "loss": 2.2589,
+      "step": 16701
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5396362480723483e-05,
+      "loss": 1.805,
+      "step": 16702
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.539585769957087e-05,
+      "loss": 1.9552,
+      "step": 16703
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5395352899021428e-05,
+      "loss": 1.8453,
+      "step": 16704
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.539484807907696e-05,
+      "loss": 1.7939,
+      "step": 16705
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5394343239739296e-05,
+      "loss": 1.7232,
+      "step": 16706
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5393838381010245e-05,
+      "loss": 1.9058,
+      "step": 16707
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.539333350289162e-05,
+      "loss": 2.5876,
+      "step": 16708
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5392828605385235e-05,
+      "loss": 1.9151,
+      "step": 16709
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5392323688492904e-05,
+      "loss": 1.8697,
+      "step": 16710
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5391818752216448e-05,
+      "loss": 1.9861,
+      "step": 16711
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5391313796557676e-05,
+      "loss": 1.8258,
+      "step": 16712
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.539080882151841e-05,
+      "loss": 1.8099,
+      "step": 16713
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.539030382710046e-05,
+      "loss": 1.749,
+      "step": 16714
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.538979881330564e-05,
+      "loss": 1.8542,
+      "step": 16715
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.538929378013577e-05,
+      "loss": 1.9919,
+      "step": 16716
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5388788727592663e-05,
+      "loss": 1.6259,
+      "step": 16717
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5388283655678135e-05,
+      "loss": 1.8211,
+      "step": 16718
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.5387778564394e-05,
+      "loss": 2.0576,
+      "step": 16719
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.538727345374208e-05,
+      "loss": 2.2135,
+      "step": 16720
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.538676832372418e-05,
+      "loss": 1.6962,
+      "step": 16721
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5386263174342126e-05,
+      "loss": 2.0064,
+      "step": 16722
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.538575800559773e-05,
+      "loss": 1.9441,
+      "step": 16723
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5385252817492804e-05,
+      "loss": 2.0318,
+      "step": 16724
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.538474761002917e-05,
+      "loss": 2.1586,
+      "step": 16725
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5384242383208638e-05,
+      "loss": 2.0976,
+      "step": 16726
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5383737137033032e-05,
+      "loss": 1.846,
+      "step": 16727
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5383231871504166e-05,
+      "loss": 1.9682,
+      "step": 16728
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5382726586623847e-05,
+      "loss": 2.1867,
+      "step": 16729
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.538222128239391e-05,
+      "loss": 1.8425,
+      "step": 16730
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.538171595881615e-05,
+      "loss": 1.8321,
+      "step": 16731
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.53812106158924e-05,
+      "loss": 2.16,
+      "step": 16732
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5380705253624467e-05,
+      "loss": 2.0902,
+      "step": 16733
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5380199872014174e-05,
+      "loss": 1.9269,
+      "step": 16734
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5379694471063335e-05,
+      "loss": 2.3709,
+      "step": 16735
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.537918905077376e-05,
+      "loss": 1.921,
+      "step": 16736
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.537868361114728e-05,
+      "loss": 1.9806,
+      "step": 16737
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5378178152185703e-05,
+      "loss": 2.063,
+      "step": 16738
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5377672673890844e-05,
+      "loss": 1.9654,
+      "step": 16739
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5377167176264526e-05,
+      "loss": 1.8495,
+      "step": 16740
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5376661659308564e-05,
+      "loss": 1.7284,
+      "step": 16741
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5376156123024776e-05,
+      "loss": 1.7545,
+      "step": 16742
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.5375650567414977e-05,
+      "loss": 2.0689,
+      "step": 16743
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5375144992480984e-05,
+      "loss": 1.8869,
+      "step": 16744
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.537463939822462e-05,
+      "loss": 1.6682,
+      "step": 16745
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5374133784647693e-05,
+      "loss": 1.8986,
+      "step": 16746
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.537362815175203e-05,
+      "loss": 2.0657,
+      "step": 16747
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5373122499539443e-05,
+      "loss": 1.9108,
+      "step": 16748
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.537261682801175e-05,
+      "loss": 2.115,
+      "step": 16749
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5372111137170774e-05,
+      "loss": 1.9744,
+      "step": 16750
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5371605427018325e-05,
+      "loss": 1.9371,
+      "step": 16751
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5371099697556226e-05,
+      "loss": 1.9524,
+      "step": 16752
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5370593948786296e-05,
+      "loss": 1.8578,
+      "step": 16753
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5370088180710347e-05,
+      "loss": 1.557,
+      "step": 16754
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5369582393330205e-05,
+      "loss": 1.7682,
+      "step": 16755
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.536907658664768e-05,
+      "loss": 1.7743,
+      "step": 16756
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5368570760664598e-05,
+      "loss": 2.0569,
+      "step": 16757
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.536806491538277e-05,
+      "loss": 2.0658,
+      "step": 16758
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5367559050804023e-05,
+      "loss": 1.9531,
+      "step": 16759
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.536705316693017e-05,
+      "loss": 1.7063,
+      "step": 16760
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5366547263763025e-05,
+      "loss": 1.7144,
+      "step": 16761
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5366041341304415e-05,
+      "loss": 1.8118,
+      "step": 16762
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5365535399556154e-05,
+      "loss": 1.5652,
+      "step": 16763
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5365029438520065e-05,
+      "loss": 1.9172,
+      "step": 16764
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5364523458197965e-05,
+      "loss": 1.5918,
+      "step": 16765
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5364017458591668e-05,
+      "loss": 1.8755,
+      "step": 16766
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5363511439703002e-05,
+      "loss": 1.8494,
+      "step": 16767
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5363005401533772e-05,
+      "loss": 1.6264,
+      "step": 16768
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.536249934408581e-05,
+      "loss": 2.1083,
+      "step": 16769
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5361993267360932e-05,
+      "loss": 2.0323,
+      "step": 16770
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.536148717136096e-05,
+      "loss": 1.7066,
+      "step": 16771
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.53609810560877e-05,
+      "loss": 2.0089,
+      "step": 16772
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.536047492154299e-05,
+      "loss": 2.2182,
+      "step": 16773
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5359968767728638e-05,
+      "loss": 1.9883,
+      "step": 16774
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5359462594646464e-05,
+      "loss": 1.9792,
+      "step": 16775
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.535895640229829e-05,
+      "loss": 1.9742,
+      "step": 16776
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.5358450190685936e-05,
+      "loss": 1.7167,
+      "step": 16777
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.5357943959811222e-05,
+      "loss": 2.6585,
+      "step": 16778
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5357437709675964e-05,
+      "loss": 1.9874,
+      "step": 16779
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5356931440281987e-05,
+      "loss": 2.2432,
+      "step": 16780
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.5356425151631108e-05,
+      "loss": 1.8715,
+      "step": 16781
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.535591884372514e-05,
+      "loss": 1.7755,
+      "step": 16782
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.535541251656592e-05,
+      "loss": 1.8858,
+      "step": 16783
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.5354906170155252e-05,
+      "loss": 2.0466,
+      "step": 16784
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5354399804494967e-05,
+      "loss": 1.7855,
+      "step": 16785
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.535389341958688e-05,
+      "loss": 1.6701,
+      "step": 16786
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5353387015432814e-05,
+      "loss": 1.9307,
+      "step": 16787
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.5352880592034585e-05,
+      "loss": 2.1359,
+      "step": 16788
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5352374149394018e-05,
+      "loss": 1.5983,
+      "step": 16789
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.5351867687512927e-05,
+      "loss": 1.6874,
+      "step": 16790
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5351361206393144e-05,
+      "loss": 1.7422,
+      "step": 16791
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5350854706036476e-05,
+      "loss": 1.7576,
+      "step": 16792
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.5350348186444758e-05,
+      "loss": 1.9908,
+      "step": 16793
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5349841647619798e-05,
+      "loss": 1.9087,
+      "step": 16794
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5349335089563425e-05,
+      "loss": 1.9059,
+      "step": 16795
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5348828512277457e-05,
+      "loss": 2.0685,
+      "step": 16796
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.534832191576372e-05,
+      "loss": 1.7981,
+      "step": 16797
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5347815300024023e-05,
+      "loss": 1.7246,
+      "step": 16798
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5347308665060197e-05,
+      "loss": 1.6177,
+      "step": 16799
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.5346802010874062e-05,
+      "loss": 2.1529,
+      "step": 16800
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.5346295337467442e-05,
+      "loss": 1.4013,
+      "step": 16801
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.534578864484215e-05,
+      "loss": 1.6847,
+      "step": 16802
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.534528193300001e-05,
+      "loss": 2.0729,
+      "step": 16803
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.534477520194285e-05,
+      "loss": 1.8792,
+      "step": 16804
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5344268451672486e-05,
+      "loss": 1.7852,
+      "step": 16805
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5343761682190737e-05,
+      "loss": 1.8867,
+      "step": 16806
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5343254893499437e-05,
+      "loss": 2.0726,
+      "step": 16807
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.534274808560039e-05,
+      "loss": 1.8727,
+      "step": 16808
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5342241258495434e-05,
+      "loss": 1.9461,
+      "step": 16809
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5341734412186377e-05,
+      "loss": 1.8434,
+      "step": 16810
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5341227546675056e-05,
+      "loss": 1.9586,
+      "step": 16811
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.534072066196328e-05,
+      "loss": 1.4322,
+      "step": 16812
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.534021375805288e-05,
+      "loss": 1.8882,
+      "step": 16813
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5339706834945673e-05,
+      "loss": 2.0536,
+      "step": 16814
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.533919989264348e-05,
+      "loss": 1.78,
+      "step": 16815
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5338692931148128e-05,
+      "loss": 1.9282,
+      "step": 16816
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5338185950461437e-05,
+      "loss": 2.0386,
+      "step": 16817
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.533767895058523e-05,
+      "loss": 1.7476,
+      "step": 16818
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5337171931521333e-05,
+      "loss": 1.8687,
+      "step": 16819
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.533666489327156e-05,
+      "loss": 1.9791,
+      "step": 16820
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5336157835837743e-05,
+      "loss": 1.9299,
+      "step": 16821
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.53356507592217e-05,
+      "loss": 1.9862,
+      "step": 16822
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.533514366342525e-05,
+      "loss": 1.5652,
+      "step": 16823
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5334636548450224e-05,
+      "loss": 1.8288,
+      "step": 16824
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5334129414298443e-05,
+      "loss": 1.7728,
+      "step": 16825
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5333622260971724e-05,
+      "loss": 1.985,
+      "step": 16826
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.533311508847189e-05,
+      "loss": 2.0729,
+      "step": 16827
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5332607896800783e-05,
+      "loss": 1.674,
+      "step": 16828
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5332100685960198e-05,
+      "loss": 2.1187,
+      "step": 16829
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.533159345595198e-05,
+      "loss": 1.6801,
+      "step": 16830
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5331086206777942e-05,
+      "loss": 1.6589,
+      "step": 16831
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.533057893843991e-05,
+      "loss": 1.5668,
+      "step": 16832
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5330071650939705e-05,
+      "loss": 1.793,
+      "step": 16833
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5329564344279155e-05,
+      "loss": 1.6703,
+      "step": 16834
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.532905701846008e-05,
+      "loss": 1.8543,
+      "step": 16835
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5328549673484307e-05,
+      "loss": 1.6874,
+      "step": 16836
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5328042309353655e-05,
+      "loss": 2.0452,
+      "step": 16837
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.532753492606996e-05,
+      "loss": 1.8185,
+      "step": 16838
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5327027523635027e-05,
+      "loss": 2.0375,
+      "step": 16839
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5326520102050693e-05,
+      "loss": 2.0428,
+      "step": 16840
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.532601266131878e-05,
+      "loss": 1.9474,
+      "step": 16841
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5325505201441112e-05,
+      "loss": 1.7443,
+      "step": 16842
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.532499772241951e-05,
+      "loss": 1.7349,
+      "step": 16843
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5324490224255803e-05,
+      "loss": 1.9196,
+      "step": 16844
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.532398270695181e-05,
+      "loss": 2.0464,
+      "step": 16845
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.532347517050936e-05,
+      "loss": 1.7556,
+      "step": 16846
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5322967614930272e-05,
+      "loss": 1.8849,
+      "step": 16847
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.532246004021638e-05,
+      "loss": 2.0856,
+      "step": 16848
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.53219524463695e-05,
+      "loss": 1.939,
+      "step": 16849
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.532144483339146e-05,
+      "loss": 1.9846,
+      "step": 16850
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5320937201284084e-05,
+      "loss": 1.9949,
+      "step": 16851
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5320429550049195e-05,
+      "loss": 1.5574,
+      "step": 16852
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.531992187968862e-05,
+      "loss": 1.7089,
+      "step": 16853
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.531941419020419e-05,
+      "loss": 2.0743,
+      "step": 16854
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.531890648159772e-05,
+      "loss": 1.9707,
+      "step": 16855
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.531839875387104e-05,
+      "loss": 1.9494,
+      "step": 16856
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5317891007025972e-05,
+      "loss": 1.9392,
+      "step": 16857
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5317383241064344e-05,
+      "loss": 1.7238,
+      "step": 16858
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.531687545598798e-05,
+      "loss": 2.015,
+      "step": 16859
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5316367651798708e-05,
+      "loss": 1.5317,
+      "step": 16860
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.5315859828498353e-05,
+      "loss": 2.1321,
+      "step": 16861
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.531535198608874e-05,
+      "loss": 1.7879,
+      "step": 16862
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.531484412457169e-05,
+      "loss": 1.5916,
+      "step": 16863
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.531433624394903e-05,
+      "loss": 1.9626,
+      "step": 16864
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5313828344222598e-05,
+      "loss": 1.9058,
+      "step": 16865
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.53133204253942e-05,
+      "loss": 1.9087,
+      "step": 16866
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5312812487465675e-05,
+      "loss": 1.9214,
+      "step": 16867
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5312304530438845e-05,
+      "loss": 1.7234,
+      "step": 16868
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.531179655431554e-05,
+      "loss": 2.0708,
+      "step": 16869
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5311288559097576e-05,
+      "loss": 1.7597,
+      "step": 16870
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.531078054478679e-05,
+      "loss": 1.9047,
+      "step": 16871
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5310272511385003e-05,
+      "loss": 2.1653,
+      "step": 16872
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.530976445889404e-05,
+      "loss": 2.0068,
+      "step": 16873
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5309256387315734e-05,
+      "loss": 1.5881,
+      "step": 16874
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5308748296651907e-05,
+      "loss": 1.7699,
+      "step": 16875
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5308240186904382e-05,
+      "loss": 1.9001,
+      "step": 16876
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.530773205807499e-05,
+      "loss": 1.7059,
+      "step": 16877
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5307223910165556e-05,
+      "loss": 2.0381,
+      "step": 16878
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.530671574317791e-05,
+      "loss": 1.6912,
+      "step": 16879
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.530620755711387e-05,
+      "loss": 2.3308,
+      "step": 16880
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5305699351975273e-05,
+      "loss": 1.7824,
+      "step": 16881
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.530519112776394e-05,
+      "loss": 1.7888,
+      "step": 16882
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5304682884481704e-05,
+      "loss": 1.9865,
+      "step": 16883
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5304174622130383e-05,
+      "loss": 1.9363,
+      "step": 16884
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5303666340711808e-05,
+      "loss": 1.6266,
+      "step": 16885
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.530315804022781e-05,
+      "loss": 1.8333,
+      "step": 16886
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5302649720680208e-05,
+      "loss": 1.8425,
+      "step": 16887
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5302141382070842e-05,
+      "loss": 2.0328,
+      "step": 16888
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5301633024401525e-05,
+      "loss": 1.8863,
+      "step": 16889
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5301124647674094e-05,
+      "loss": 2.0311,
+      "step": 16890
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5300616251890374e-05,
+      "loss": 1.7508,
+      "step": 16891
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.530010783705219e-05,
+      "loss": 2.0296,
+      "step": 16892
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5299599403161373e-05,
+      "loss": 1.6105,
+      "step": 16893
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5299090950219748e-05,
+      "loss": 1.7978,
+      "step": 16894
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5298582478229148e-05,
+      "loss": 2.0241,
+      "step": 16895
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5298073987191395e-05,
+      "loss": 1.8023,
+      "step": 16896
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5297565477108318e-05,
+      "loss": 2.0477,
+      "step": 16897
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5297056947981747e-05,
+      "loss": 1.8162,
+      "step": 16898
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5296548399813506e-05,
+      "loss": 2.1422,
+      "step": 16899
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5296039832605432e-05,
+      "loss": 1.6841,
+      "step": 16900
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5295531246359344e-05,
+      "loss": 2.002,
+      "step": 16901
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5295022641077074e-05,
+      "loss": 2.0369,
+      "step": 16902
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.529451401676045e-05,
+      "loss": 1.9768,
+      "step": 16903
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5294005373411297e-05,
+      "loss": 1.744,
+      "step": 16904
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.529349671103145e-05,
+      "loss": 2.171,
+      "step": 16905
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5292988029622732e-05,
+      "loss": 2.0173,
+      "step": 16906
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5292479329186977e-05,
+      "loss": 1.8173,
+      "step": 16907
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5291970609726008e-05,
+      "loss": 2.0719,
+      "step": 16908
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5291461871241655e-05,
+      "loss": 1.6686,
+      "step": 16909
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.529095311373575e-05,
+      "loss": 1.7749,
+      "step": 16910
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.529044433721012e-05,
+      "loss": 3.7265,
+      "step": 16911
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5289935541666595e-05,
+      "loss": 1.9015,
+      "step": 16912
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5289426727107e-05,
+      "loss": 1.5011,
+      "step": 16913
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5288917893533167e-05,
+      "loss": 2.0223,
+      "step": 16914
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5288409040946927e-05,
+      "loss": 1.648,
+      "step": 16915
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5287900169350106e-05,
+      "loss": 1.9413,
+      "step": 16916
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5287391278744537e-05,
+      "loss": 1.8738,
+      "step": 16917
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5286882369132044e-05,
+      "loss": 1.8356,
+      "step": 16918
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.528637344051446e-05,
+      "loss": 2.0716,
+      "step": 16919
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5285864492893612e-05,
+      "loss": 1.8902,
+      "step": 16920
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5285355526271333e-05,
+      "loss": 1.8836,
+      "step": 16921
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5284846540649447e-05,
+      "loss": 1.9748,
+      "step": 16922
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5284337536029795e-05,
+      "loss": 1.8577,
+      "step": 16923
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5283828512414195e-05,
+      "loss": 1.9614,
+      "step": 16924
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.528331946980448e-05,
+      "loss": 2.1122,
+      "step": 16925
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5282810408202483e-05,
+      "loss": 1.9159,
+      "step": 16926
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.528230132761003e-05,
+      "loss": 1.7786,
+      "step": 16927
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5281792228028955e-05,
+      "loss": 1.8711,
+      "step": 16928
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.528128310946108e-05,
+      "loss": 1.9893,
+      "step": 16929
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.528077397190825e-05,
+      "loss": 1.9774,
+      "step": 16930
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.528026481537228e-05,
+      "loss": 1.9277,
+      "step": 16931
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5279755639855013e-05,
+      "loss": 1.8342,
+      "step": 16932
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.527924644535827e-05,
+      "loss": 2.1418,
+      "step": 16933
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.527873723188388e-05,
+      "loss": 2.0736,
+      "step": 16934
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5278227999433686e-05,
+      "loss": 2.1134,
+      "step": 16935
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5277718748009504e-05,
+      "loss": 1.8191,
+      "step": 16936
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5277209477613176e-05,
+      "loss": 1.8253,
+      "step": 16937
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5276700188246526e-05,
+      "loss": 1.8306,
+      "step": 16938
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5276190879911385e-05,
+      "loss": 2.158,
+      "step": 16939
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.527568155260959e-05,
+      "loss": 1.7527,
+      "step": 16940
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.5275172206342965e-05,
+      "loss": 1.9587,
+      "step": 16941
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.527466284111334e-05,
+      "loss": 1.7585,
+      "step": 16942
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5274153456922554e-05,
+      "loss": 1.7659,
+      "step": 16943
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5273644053772432e-05,
+      "loss": 1.8556,
+      "step": 16944
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.527313463166481e-05,
+      "loss": 1.8489,
+      "step": 16945
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.527262519060151e-05,
+      "loss": 1.7461,
+      "step": 16946
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5272115730584372e-05,
+      "loss": 1.727,
+      "step": 16947
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5271606251615228e-05,
+      "loss": 1.54,
+      "step": 16948
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5271096753695906e-05,
+      "loss": 1.7373,
+      "step": 16949
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.527058723682823e-05,
+      "loss": 1.7207,
+      "step": 16950
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5270077701014044e-05,
+      "loss": 1.9778,
+      "step": 16951
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5269568146255176e-05,
+      "loss": 1.9422,
+      "step": 16952
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5269058572553457e-05,
+      "loss": 1.6189,
+      "step": 16953
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.526854897991072e-05,
+      "loss": 1.8963,
+      "step": 16954
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5268039368328794e-05,
+      "loss": 1.9711,
+      "step": 16955
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.526752973780951e-05,
+      "loss": 1.7699,
+      "step": 16956
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5267020088354704e-05,
+      "loss": 2.0039,
+      "step": 16957
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5266510419966204e-05,
+      "loss": 2.1157,
+      "step": 16958
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.526600073264585e-05,
+      "loss": 2.0729,
+      "step": 16959
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5265491026395464e-05,
+      "loss": 1.9359,
+      "step": 16960
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5264981301216885e-05,
+      "loss": 2.2167,
+      "step": 16961
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5264471557111943e-05,
+      "loss": 1.6938,
+      "step": 16962
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5263961794082473e-05,
+      "loss": 1.734,
+      "step": 16963
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.52634520121303e-05,
+      "loss": 1.7675,
+      "step": 16964
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.526294221125727e-05,
+      "loss": 1.9532,
+      "step": 16965
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.375,
+      "learning_rate": 1.52624323914652e-05,
+      "loss": 1.6392,
+      "step": 16966
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.526192255275593e-05,
+      "loss": 1.6927,
+      "step": 16967
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5261412695131295e-05,
+      "loss": 2.2571,
+      "step": 16968
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5260902818593125e-05,
+      "loss": 1.8759,
+      "step": 16969
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5260392923143257e-05,
+      "loss": 1.8003,
+      "step": 16970
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5259883008783514e-05,
+      "loss": 1.7603,
+      "step": 16971
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5259373075515743e-05,
+      "loss": 1.8635,
+      "step": 16972
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5258863123341766e-05,
+      "loss": 2.1836,
+      "step": 16973
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.525835315226342e-05,
+      "loss": 1.6997,
+      "step": 16974
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5257843162282537e-05,
+      "loss": 2.0011,
+      "step": 16975
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.525733315340095e-05,
+      "loss": 2.173,
+      "step": 16976
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5256823125620497e-05,
+      "loss": 1.5096,
+      "step": 16977
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5256313078943005e-05,
+      "loss": 1.7593,
+      "step": 16978
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5255803013370314e-05,
+      "loss": 1.974,
+      "step": 16979
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5255292928904253e-05,
+      "loss": 1.8851,
+      "step": 16980
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5254782825546653e-05,
+      "loss": 1.9371,
+      "step": 16981
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5254272703299358e-05,
+      "loss": 1.8027,
+      "step": 16982
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5253762562164189e-05,
+      "loss": 2.0484,
+      "step": 16983
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5253252402142989e-05,
+      "loss": 1.7541,
+      "step": 16984
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.525274222323759e-05,
+      "loss": 1.7972,
+      "step": 16985
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5252232025449822e-05,
+      "loss": 2.1505,
+      "step": 16986
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5251721808781526e-05,
+      "loss": 2.0812,
+      "step": 16987
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.5251211573234528e-05,
+      "loss": 1.7224,
+      "step": 16988
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.375,
+      "learning_rate": 1.525070131881067e-05,
+      "loss": 1.6264,
+      "step": 16989
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.525019104551178e-05,
+      "loss": 1.8162,
+      "step": 16990
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5249680753339695e-05,
+      "loss": 1.6425,
+      "step": 16991
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5249170442296249e-05,
+      "loss": 1.9548,
+      "step": 16992
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.524866011238328e-05,
+      "loss": 2.1201,
+      "step": 16993
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5248149763602616e-05,
+      "loss": 1.6719,
+      "step": 16994
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5247639395956093e-05,
+      "loss": 1.7914,
+      "step": 16995
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5247129009445552e-05,
+      "loss": 1.9885,
+      "step": 16996
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.5246618604072819e-05,
+      "loss": 1.523,
+      "step": 16997
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5246108179839736e-05,
+      "loss": 1.6016,
+      "step": 16998
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5245597736748133e-05,
+      "loss": 1.6888,
+      "step": 16999
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5245087274799848e-05,
+      "loss": 1.6644,
+      "step": 17000
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5244576793996712e-05,
+      "loss": 1.9644,
+      "step": 17001
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5244066294340565e-05,
+      "loss": 2.0426,
+      "step": 17002
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.524355577583324e-05,
+      "loss": 1.9439,
+      "step": 17003
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.524304523847657e-05,
+      "loss": 1.7056,
+      "step": 17004
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5242534682272394e-05,
+      "loss": 1.8169,
+      "step": 17005
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5242024107222545e-05,
+      "loss": 1.6731,
+      "step": 17006
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.5241513513328861e-05,
+      "loss": 1.9878,
+      "step": 17007
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5241002900593174e-05,
+      "loss": 2.0928,
+      "step": 17008
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5240492269017319e-05,
+      "loss": 1.8481,
+      "step": 17009
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5239981618603136e-05,
+      "loss": 2.3804,
+      "step": 17010
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5239470949352457e-05,
+      "loss": 1.6443,
+      "step": 17011
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5238960261267121e-05,
+      "loss": 1.8745,
+      "step": 17012
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5238449554348961e-05,
+      "loss": 1.7564,
+      "step": 17013
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5237938828599812e-05,
+      "loss": 1.5025,
+      "step": 17014
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5237428084021515e-05,
+      "loss": 1.93,
+      "step": 17015
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.52369173206159e-05,
+      "loss": 2.0598,
+      "step": 17016
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5236406538384807e-05,
+      "loss": 1.934,
+      "step": 17017
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5235895737330069e-05,
+      "loss": 1.9041,
+      "step": 17018
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5235384917453528e-05,
+      "loss": 2.0673,
+      "step": 17019
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5234874078757011e-05,
+      "loss": 1.6961,
+      "step": 17020
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5234363221242365e-05,
+      "loss": 1.9416,
+      "step": 17021
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5233852344911417e-05,
+      "loss": 1.8247,
+      "step": 17022
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.523334144976601e-05,
+      "loss": 2.1743,
+      "step": 17023
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5232830535807978e-05,
+      "loss": 1.7597,
+      "step": 17024
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.5232319603039156e-05,
+      "loss": 2.0583,
+      "step": 17025
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5231808651461384e-05,
+      "loss": 2.3481,
+      "step": 17026
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5231297681076495e-05,
+      "loss": 1.6769,
+      "step": 17027
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.523078669188633e-05,
+      "loss": 1.7225,
+      "step": 17028
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5230275683892725e-05,
+      "loss": 1.9833,
+      "step": 17029
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5229764657097514e-05,
+      "loss": 2.2635,
+      "step": 17030
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5229253611502537e-05,
+      "loss": 1.8785,
+      "step": 17031
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.522874254710963e-05,
+      "loss": 2.0752,
+      "step": 17032
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5228231463920629e-05,
+      "loss": 1.8172,
+      "step": 17033
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5227720361937373e-05,
+      "loss": 2.0406,
+      "step": 17034
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5227209241161695e-05,
+      "loss": 1.8206,
+      "step": 17035
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5226698101595441e-05,
+      "loss": 1.9809,
+      "step": 17036
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5226186943240442e-05,
+      "loss": 2.0604,
+      "step": 17037
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5225675766098538e-05,
+      "loss": 1.9112,
+      "step": 17038
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5225164570171563e-05,
+      "loss": 1.9499,
+      "step": 17039
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.5224653355461359e-05,
+      "loss": 1.9959,
+      "step": 17040
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.522414212196976e-05,
+      "loss": 1.9889,
+      "step": 17041
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5223630869698607e-05,
+      "loss": 1.9959,
+      "step": 17042
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5223119598649736e-05,
+      "loss": 1.8745,
+      "step": 17043
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5222608308824984e-05,
+      "loss": 1.6951,
+      "step": 17044
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5222097000226193e-05,
+      "loss": 1.637,
+      "step": 17045
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5221585672855196e-05,
+      "loss": 2.0706,
+      "step": 17046
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5221074326713835e-05,
+      "loss": 1.9563,
+      "step": 17047
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.5220562961803942e-05,
+      "loss": 2.1806,
+      "step": 17048
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5220051578127369e-05,
+      "loss": 2.0909,
+      "step": 17049
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5219540175685938e-05,
+      "loss": 1.9613,
+      "step": 17050
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5219028754481495e-05,
+      "loss": 2.0788,
+      "step": 17051
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5218517314515878e-05,
+      "loss": 1.6514,
+      "step": 17052
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5218005855790927e-05,
+      "loss": 1.8856,
+      "step": 17053
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5217494378308477e-05,
+      "loss": 1.9129,
+      "step": 17054
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5216982882070368e-05,
+      "loss": 1.7445,
+      "step": 17055
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5216471367078444e-05,
+      "loss": 2.0715,
+      "step": 17056
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5215959833334535e-05,
+      "loss": 1.9556,
+      "step": 17057
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5215448280840485e-05,
+      "loss": 1.7286,
+      "step": 17058
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5214936709598134e-05,
+      "loss": 1.9079,
+      "step": 17059
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5214425119609315e-05,
+      "loss": 1.737,
+      "step": 17060
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.5213913510875873e-05,
+      "loss": 1.9434,
+      "step": 17061
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5213401883399643e-05,
+      "loss": 1.879,
+      "step": 17062
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.521289023718247e-05,
+      "loss": 1.7427,
+      "step": 17063
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5212378572226187e-05,
+      "loss": 2.0285,
+      "step": 17064
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5211866888532636e-05,
+      "loss": 2.4227,
+      "step": 17065
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5211355186103655e-05,
+      "loss": 1.6568,
+      "step": 17066
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5210843464941087e-05,
+      "loss": 1.8263,
+      "step": 17067
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5210331725046768e-05,
+      "loss": 1.9634,
+      "step": 17068
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5209819966422536e-05,
+      "loss": 1.5603,
+      "step": 17069
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5209308189070238e-05,
+      "loss": 1.7989,
+      "step": 17070
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5208796392991705e-05,
+      "loss": 1.713,
+      "step": 17071
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5208284578188784e-05,
+      "loss": 1.9493,
+      "step": 17072
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5207772744663309e-05,
+      "loss": 1.9283,
+      "step": 17073
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5207260892417123e-05,
+      "loss": 1.8308,
+      "step": 17074
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5206749021452067e-05,
+      "loss": 2.0528,
+      "step": 17075
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5206237131769981e-05,
+      "loss": 1.5991,
+      "step": 17076
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5205725223372701e-05,
+      "loss": 1.868,
+      "step": 17077
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5205213296262069e-05,
+      "loss": 1.9922,
+      "step": 17078
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.520470135043993e-05,
+      "loss": 1.6659,
+      "step": 17079
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5204189385908117e-05,
+      "loss": 1.8587,
+      "step": 17080
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5203677402668472e-05,
+      "loss": 1.8612,
+      "step": 17081
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5203165400722844e-05,
+      "loss": 2.0489,
+      "step": 17082
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5202653380073064e-05,
+      "loss": 1.825,
+      "step": 17083
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5202141340720975e-05,
+      "loss": 2.1338,
+      "step": 17084
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5201629282668418e-05,
+      "loss": 1.6967,
+      "step": 17085
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5201117205917232e-05,
+      "loss": 1.7668,
+      "step": 17086
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5200605110469265e-05,
+      "loss": 1.71,
+      "step": 17087
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5200092996326348e-05,
+      "loss": 1.9078,
+      "step": 17088
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5199580863490328e-05,
+      "loss": 2.0555,
+      "step": 17089
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5199068711963043e-05,
+      "loss": 1.6018,
+      "step": 17090
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.5198556541746339e-05,
+      "loss": 1.5676,
+      "step": 17091
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.519804435284205e-05,
+      "loss": 2.0012,
+      "step": 17092
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5197532145252023e-05,
+      "loss": 2.0198,
+      "step": 17093
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5197019918978097e-05,
+      "loss": 1.9734,
+      "step": 17094
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5196507674022112e-05,
+      "loss": 2.1687,
+      "step": 17095
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5195995410385912e-05,
+      "loss": 2.0507,
+      "step": 17096
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5195483128071335e-05,
+      "loss": 1.8193,
+      "step": 17097
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5194970827080226e-05,
+      "loss": 1.8845,
+      "step": 17098
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.519445850741443e-05,
+      "loss": 1.8348,
+      "step": 17099
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5193946169075777e-05,
+      "loss": 1.603,
+      "step": 17100
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5193433812066122e-05,
+      "loss": 1.7723,
+      "step": 17101
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.5192921436387296e-05,
+      "loss": 2.5228,
+      "step": 17102
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5192409042041149e-05,
+      "loss": 1.8601,
+      "step": 17103
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.5191896629029516e-05,
+      "loss": 2.1095,
+      "step": 17104
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5191384197354246e-05,
+      "loss": 2.1655,
+      "step": 17105
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5190871747017175e-05,
+      "loss": 1.8912,
+      "step": 17106
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5190359278020151e-05,
+      "loss": 2.0453,
+      "step": 17107
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5189846790365009e-05,
+      "loss": 2.1827,
+      "step": 17108
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5189334284053593e-05,
+      "loss": 2.0422,
+      "step": 17109
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5188821759087754e-05,
+      "loss": 1.7314,
+      "step": 17110
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5188309215469325e-05,
+      "loss": 1.8382,
+      "step": 17111
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5187796653200151e-05,
+      "loss": 1.6337,
+      "step": 17112
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5187284072282076e-05,
+      "loss": 2.1299,
+      "step": 17113
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5186771472716939e-05,
+      "loss": 2.1454,
+      "step": 17114
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5186258854506587e-05,
+      "loss": 1.8354,
+      "step": 17115
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5185746217652861e-05,
+      "loss": 1.6462,
+      "step": 17116
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5185233562157607e-05,
+      "loss": 2.0459,
+      "step": 17117
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5184720888022657e-05,
+      "loss": 1.9726,
+      "step": 17118
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5184208195249868e-05,
+      "loss": 1.9195,
+      "step": 17119
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5183695483841074e-05,
+      "loss": 1.7175,
+      "step": 17120
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5183182753798123e-05,
+      "loss": 1.5385,
+      "step": 17121
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5182670005122853e-05,
+      "loss": 2.053,
+      "step": 17122
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5182157237817113e-05,
+      "loss": 1.8932,
+      "step": 17123
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.518164445188274e-05,
+      "loss": 1.8707,
+      "step": 17124
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.5181131647321578e-05,
+      "loss": 2.0477,
+      "step": 17125
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5180618824135477e-05,
+      "loss": 1.8595,
+      "step": 17126
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5180105982326278e-05,
+      "loss": 2.1434,
+      "step": 17127
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5179593121895822e-05,
+      "loss": 1.7712,
+      "step": 17128
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5179080242845952e-05,
+      "loss": 1.9799,
+      "step": 17129
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5178567345178517e-05,
+      "loss": 2.0071,
+      "step": 17130
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.5178054428895355e-05,
+      "loss": 1.7968,
+      "step": 17131
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.517754149399831e-05,
+      "loss": 2.1905,
+      "step": 17132
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5177028540489229e-05,
+      "loss": 1.7611,
+      "step": 17133
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5176515568369957e-05,
+      "loss": 1.8479,
+      "step": 17134
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.517600257764233e-05,
+      "loss": 2.0508,
+      "step": 17135
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5175489568308201e-05,
+      "loss": 1.9046,
+      "step": 17136
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5174976540369413e-05,
+      "loss": 1.6551,
+      "step": 17137
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.5174463493827807e-05,
+      "loss": 2.1676,
+      "step": 17138
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5173950428685227e-05,
+      "loss": 1.9265,
+      "step": 17139
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.517343734494352e-05,
+      "loss": 1.5255,
+      "step": 17140
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.517292424260453e-05,
+      "loss": 1.8773,
+      "step": 17141
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5172411121670099e-05,
+      "loss": 2.0257,
+      "step": 17142
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.5171897982142075e-05,
+      "loss": 1.5726,
+      "step": 17143
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5171384824022299e-05,
+      "loss": 1.8349,
+      "step": 17144
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5170871647312617e-05,
+      "loss": 1.9379,
+      "step": 17145
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5170358452014875e-05,
+      "loss": 2.0496,
+      "step": 17146
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.516984523813092e-05,
+      "loss": 1.8043,
+      "step": 17147
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5169332005662591e-05,
+      "loss": 1.9566,
+      "step": 17148
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5168818754611736e-05,
+      "loss": 1.9721,
+      "step": 17149
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5168305484980204e-05,
+      "loss": 1.7736,
+      "step": 17150
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.516779219676983e-05,
+      "loss": 1.6136,
+      "step": 17151
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.516727888998247e-05,
+      "loss": 1.6731,
+      "step": 17152
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5166765564619961e-05,
+      "loss": 1.7001,
+      "step": 17153
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5166252220684154e-05,
+      "loss": 1.8211,
+      "step": 17154
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5165738858176892e-05,
+      "loss": 2.0286,
+      "step": 17155
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5165225477100019e-05,
+      "loss": 1.7166,
+      "step": 17156
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.5164712077455383e-05,
+      "loss": 1.9279,
+      "step": 17157
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5164198659244829e-05,
+      "loss": 2.1146,
+      "step": 17158
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.51636852224702e-05,
+      "loss": 1.9074,
+      "step": 17159
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5163171767133347e-05,
+      "loss": 1.8783,
+      "step": 17160
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.516265829323611e-05,
+      "loss": 2.0124,
+      "step": 17161
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5162144800780336e-05,
+      "loss": 1.8282,
+      "step": 17162
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5161631289767878e-05,
+      "loss": 1.6719,
+      "step": 17163
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5161117760200572e-05,
+      "loss": 1.7991,
+      "step": 17164
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5160604212080265e-05,
+      "loss": 2.0733,
+      "step": 17165
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.5160090645408811e-05,
+      "loss": 1.308,
+      "step": 17166
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.515957706018805e-05,
+      "loss": 1.8925,
+      "step": 17167
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5159063456419832e-05,
+      "loss": 1.6132,
+      "step": 17168
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5158549834105998e-05,
+      "loss": 2.0861,
+      "step": 17169
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.51580361932484e-05,
+      "loss": 1.6702,
+      "step": 17170
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5157522533848878e-05,
+      "loss": 1.7008,
+      "step": 17171
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5157008855909286e-05,
+      "loss": 1.8045,
+      "step": 17172
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5156495159431463e-05,
+      "loss": 1.6863,
+      "step": 17173
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5155981444417262e-05,
+      "loss": 1.7565,
+      "step": 17174
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5155467710868527e-05,
+      "loss": 1.9152,
+      "step": 17175
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5154953958787104e-05,
+      "loss": 1.669,
+      "step": 17176
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5154440188174842e-05,
+      "loss": 1.7332,
+      "step": 17177
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5153926399033588e-05,
+      "loss": 2.2527,
+      "step": 17178
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.515341259136518e-05,
+      "loss": 1.5943,
+      "step": 17179
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.515289876517148e-05,
+      "loss": 1.6123,
+      "step": 17180
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5152384920454324e-05,
+      "loss": 1.713,
+      "step": 17181
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5151871057215563e-05,
+      "loss": 1.6878,
+      "step": 17182
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5151357175457047e-05,
+      "loss": 2.0355,
+      "step": 17183
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5150843275180615e-05,
+      "loss": 1.9938,
+      "step": 17184
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.5150329356388126e-05,
+      "loss": 2.5553,
+      "step": 17185
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5149815419081412e-05,
+      "loss": 1.9716,
+      "step": 17186
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.514930146326234e-05,
+      "loss": 1.8111,
+      "step": 17187
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5148787488932741e-05,
+      "loss": 1.6945,
+      "step": 17188
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5148273496094468e-05,
+      "loss": 2.0764,
+      "step": 17189
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5147759484749372e-05,
+      "loss": 2.0854,
+      "step": 17190
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5147245454899296e-05,
+      "loss": 1.753,
+      "step": 17191
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.514673140654609e-05,
+      "loss": 2.0407,
+      "step": 17192
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5146217339691602e-05,
+      "loss": 1.6895,
+      "step": 17193
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5145703254337685e-05,
+      "loss": 2.0335,
+      "step": 17194
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5145189150486172e-05,
+      "loss": 1.899,
+      "step": 17195
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5144675028138927e-05,
+      "loss": 1.8516,
+      "step": 17196
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5144160887297792e-05,
+      "loss": 2.0533,
+      "step": 17197
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5143646727964614e-05,
+      "loss": 1.8746,
+      "step": 17198
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5143132550141242e-05,
+      "loss": 2.0184,
+      "step": 17199
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.5142618353829525e-05,
+      "loss": 1.7473,
+      "step": 17200
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5142104139031313e-05,
+      "loss": 2.0104,
+      "step": 17201
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5141589905748449e-05,
+      "loss": 1.7621,
+      "step": 17202
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5141075653982786e-05,
+      "loss": 1.5458,
+      "step": 17203
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5140561383736175e-05,
+      "loss": 1.7637,
+      "step": 17204
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.514004709501046e-05,
+      "loss": 1.8486,
+      "step": 17205
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5139532787807491e-05,
+      "loss": 2.0295,
+      "step": 17206
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.5139018462129115e-05,
+      "loss": 1.7795,
+      "step": 17207
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5138504117977188e-05,
+      "loss": 1.857,
+      "step": 17208
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5137989755353548e-05,
+      "loss": 1.7843,
+      "step": 17209
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5137475374260053e-05,
+      "loss": 1.5891,
+      "step": 17210
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5136960974698554e-05,
+      "loss": 1.6821,
+      "step": 17211
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5136446556670885e-05,
+      "loss": 2.0142,
+      "step": 17212
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5135932120178911e-05,
+      "loss": 1.9354,
+      "step": 17213
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5135417665224476e-05,
+      "loss": 1.6633,
+      "step": 17214
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5134903191809426e-05,
+      "loss": 1.9362,
+      "step": 17215
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5134388699935616e-05,
+      "loss": 2.1249,
+      "step": 17216
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5133874189604893e-05,
+      "loss": 1.6013,
+      "step": 17217
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5133359660819108e-05,
+      "loss": 1.8985,
+      "step": 17218
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5132845113580103e-05,
+      "loss": 1.7283,
+      "step": 17219
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.375,
+      "learning_rate": 1.513233054788974e-05,
+      "loss": 1.8843,
+      "step": 17220
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.5131815963749858e-05,
+      "loss": 2.0018,
+      "step": 17221
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5131301361162316e-05,
+      "loss": 2.3851,
+      "step": 17222
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.5130786740128954e-05,
+      "loss": 1.9783,
+      "step": 17223
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.513027210065163e-05,
+      "loss": 1.7401,
+      "step": 17224
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5129757442732191e-05,
+      "loss": 1.8196,
+      "step": 17225
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5129242766372487e-05,
+      "loss": 1.9878,
+      "step": 17226
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5128728071574368e-05,
+      "loss": 1.9088,
+      "step": 17227
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5128213358339683e-05,
+      "loss": 1.8253,
+      "step": 17228
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.375,
+      "learning_rate": 1.5127698626670286e-05,
+      "loss": 2.0165,
+      "step": 17229
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5127183876568024e-05,
+      "loss": 1.8392,
+      "step": 17230
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5126669108034748e-05,
+      "loss": 2.0658,
+      "step": 17231
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.5126154321072311e-05,
+      "loss": 2.0077,
+      "step": 17232
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.512563951568256e-05,
+      "loss": 2.1188,
+      "step": 17233
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5125124691867347e-05,
+      "loss": 1.9758,
+      "step": 17234
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5124609849628523e-05,
+      "loss": 2.1081,
+      "step": 17235
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5124094988967939e-05,
+      "loss": 1.5653,
+      "step": 17236
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.5123580109887445e-05,
+      "loss": 2.1894,
+      "step": 17237
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5123065212388892e-05,
+      "loss": 2.0194,
+      "step": 17238
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5122550296474132e-05,
+      "loss": 1.9281,
+      "step": 17239
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5122035362145014e-05,
+      "loss": 1.8279,
+      "step": 17240
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5121520409403392e-05,
+      "loss": 1.6408,
+      "step": 17241
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.512100543825111e-05,
+      "loss": 2.0688,
+      "step": 17242
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.512049044869003e-05,
+      "loss": 1.7589,
+      "step": 17243
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5119975440721997e-05,
+      "loss": 1.8779,
+      "step": 17244
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5119460414348864e-05,
+      "loss": 1.9495,
+      "step": 17245
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5118945369572479e-05,
+      "loss": 1.798,
+      "step": 17246
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5118430306394697e-05,
+      "loss": 1.9757,
+      "step": 17247
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5117915224817368e-05,
+      "loss": 2.1614,
+      "step": 17248
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5117400124842343e-05,
+      "loss": 2.0152,
+      "step": 17249
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5116885006471477e-05,
+      "loss": 2.1515,
+      "step": 17250
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.511636986970662e-05,
+      "loss": 1.7789,
+      "step": 17251
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.511585471454962e-05,
+      "loss": 2.0412,
+      "step": 17252
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5115339541002336e-05,
+      "loss": 2.211,
+      "step": 17253
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5114824349066615e-05,
+      "loss": 2.1473,
+      "step": 17254
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.511430913874431e-05,
+      "loss": 2.0868,
+      "step": 17255
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.511379391003727e-05,
+      "loss": 1.8024,
+      "step": 17256
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5113278662947353e-05,
+      "loss": 2.0858,
+      "step": 17257
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5112763397476408e-05,
+      "loss": 1.9917,
+      "step": 17258
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.511224811362629e-05,
+      "loss": 1.7827,
+      "step": 17259
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5111732811398848e-05,
+      "loss": 1.9584,
+      "step": 17260
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.5111217490795933e-05,
+      "loss": 2.1241,
+      "step": 17261
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5110702151819403e-05,
+      "loss": 1.9314,
+      "step": 17262
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5110186794471105e-05,
+      "loss": 1.776,
+      "step": 17263
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5109671418752895e-05,
+      "loss": 1.8155,
+      "step": 17264
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5109156024666625e-05,
+      "loss": 1.9696,
+      "step": 17265
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.5108640612214144e-05,
+      "loss": 2.1513,
+      "step": 17266
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5108125181397314e-05,
+      "loss": 2.2667,
+      "step": 17267
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5107609732217979e-05,
+      "loss": 1.9482,
+      "step": 17268
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5107094264677996e-05,
+      "loss": 1.9293,
+      "step": 17269
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5106578778779212e-05,
+      "loss": 1.5772,
+      "step": 17270
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5106063274523493e-05,
+      "loss": 1.8452,
+      "step": 17271
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5105547751912677e-05,
+      "loss": 1.8108,
+      "step": 17272
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5105032210948629e-05,
+      "loss": 1.8212,
+      "step": 17273
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5104516651633195e-05,
+      "loss": 2.1079,
+      "step": 17274
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5104001073968234e-05,
+      "loss": 1.8728,
+      "step": 17275
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5103485477955592e-05,
+      "loss": 1.7346,
+      "step": 17276
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5102969863597125e-05,
+      "loss": 1.7653,
+      "step": 17277
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5102454230894697e-05,
+      "loss": 2.0261,
+      "step": 17278
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5101938579850144e-05,
+      "loss": 1.9547,
+      "step": 17279
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5101422910465333e-05,
+      "loss": 2.1786,
+      "step": 17280
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.5100907222742112e-05,
+      "loss": 1.6213,
+      "step": 17281
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5100391516682336e-05,
+      "loss": 1.971,
+      "step": 17282
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5099875792287857e-05,
+      "loss": 1.7096,
+      "step": 17283
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5099360049560532e-05,
+      "loss": 1.6231,
+      "step": 17284
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5098844288502214e-05,
+      "loss": 1.9801,
+      "step": 17285
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5098328509114754e-05,
+      "loss": 1.9238,
+      "step": 17286
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5097812711400014e-05,
+      "loss": 1.8552,
+      "step": 17287
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5097296895359839e-05,
+      "loss": 1.6783,
+      "step": 17288
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5096781060996088e-05,
+      "loss": 2.044,
+      "step": 17289
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5096265208310614e-05,
+      "loss": 1.8962,
+      "step": 17290
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5095749337305275e-05,
+      "loss": 1.5984,
+      "step": 17291
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5095233447981917e-05,
+      "loss": 1.9081,
+      "step": 17292
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.5094717540342401e-05,
+      "loss": 1.8771,
+      "step": 17293
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5094201614388582e-05,
+      "loss": 1.8395,
+      "step": 17294
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.5093685670122313e-05,
+      "loss": 2.0397,
+      "step": 17295
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5093169707545447e-05,
+      "loss": 1.9866,
+      "step": 17296
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5092653726659843e-05,
+      "loss": 1.8768,
+      "step": 17297
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.509213772746735e-05,
+      "loss": 1.7891,
+      "step": 17298
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.5091621709969829e-05,
+      "loss": 2.3846,
+      "step": 17299
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5091105674169129e-05,
+      "loss": 1.8223,
+      "step": 17300
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5090589620067114e-05,
+      "loss": 1.9997,
+      "step": 17301
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.509007354766563e-05,
+      "loss": 1.8669,
+      "step": 17302
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5089557456966533e-05,
+      "loss": 1.8782,
+      "step": 17303
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.508904134797168e-05,
+      "loss": 1.8836,
+      "step": 17304
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.508852522068293e-05,
+      "loss": 2.0654,
+      "step": 17305
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5088009075102135e-05,
+      "loss": 2.1932,
+      "step": 17306
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.508749291123115e-05,
+      "loss": 1.7954,
+      "step": 17307
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5086976729071831e-05,
+      "loss": 1.7408,
+      "step": 17308
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.5086460528626035e-05,
+      "loss": 1.8588,
+      "step": 17309
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5085944309895612e-05,
+      "loss": 1.7226,
+      "step": 17310
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.508542807288243e-05,
+      "loss": 1.9377,
+      "step": 17311
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5084911817588327e-05,
+      "loss": 2.1123,
+      "step": 17312
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5084395544015173e-05,
+      "loss": 1.9955,
+      "step": 17313
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.5083879252164821e-05,
+      "loss": 2.0978,
+      "step": 17314
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5083362942039125e-05,
+      "loss": 2.0663,
+      "step": 17315
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.508284661363994e-05,
+      "loss": 1.7451,
+      "step": 17316
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.5082330266969125e-05,
+      "loss": 2.4371,
+      "step": 17317
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5081813902028534e-05,
+      "loss": 1.8166,
+      "step": 17318
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5081297518820024e-05,
+      "loss": 1.8001,
+      "step": 17319
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.508078111734545e-05,
+      "loss": 2.4868,
+      "step": 17320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.508026469760667e-05,
+      "loss": 2.0082,
+      "step": 17321
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.507974825960554e-05,
+      "loss": 1.7592,
+      "step": 17322
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5079231803343917e-05,
+      "loss": 1.6018,
+      "step": 17323
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5078715328823655e-05,
+      "loss": 2.0149,
+      "step": 17324
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5078198836046615e-05,
+      "loss": 2.083,
+      "step": 17325
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5077682325014648e-05,
+      "loss": 1.8201,
+      "step": 17326
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.5077165795729616e-05,
+      "loss": 1.824,
+      "step": 17327
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5076649248193374e-05,
+      "loss": 1.8245,
+      "step": 17328
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5076132682407778e-05,
+      "loss": 1.8773,
+      "step": 17329
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5075616098374687e-05,
+      "loss": 1.9168,
+      "step": 17330
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5075099496095953e-05,
+      "loss": 1.7805,
+      "step": 17331
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.507458287557344e-05,
+      "loss": 1.7602,
+      "step": 17332
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5074066236809e-05,
+      "loss": 1.6996,
+      "step": 17333
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5073549579804493e-05,
+      "loss": 1.9086,
+      "step": 17334
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5073032904561774e-05,
+      "loss": 1.8635,
+      "step": 17335
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.50725162110827e-05,
+      "loss": 1.8274,
+      "step": 17336
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5071999499369133e-05,
+      "loss": 1.9575,
+      "step": 17337
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5071482769422925e-05,
+      "loss": 1.7633,
+      "step": 17338
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5070966021245937e-05,
+      "loss": 1.9253,
+      "step": 17339
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.5070449254840023e-05,
+      "loss": 1.7273,
+      "step": 17340
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5069932470207044e-05,
+      "loss": 1.7284,
+      "step": 17341
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.506941566734886e-05,
+      "loss": 2.2415,
+      "step": 17342
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5068898846267321e-05,
+      "loss": 1.8368,
+      "step": 17343
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5068382006964293e-05,
+      "loss": 1.6,
+      "step": 17344
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.506786514944163e-05,
+      "loss": 1.8199,
+      "step": 17345
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.506734827370119e-05,
+      "loss": 1.7311,
+      "step": 17346
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5066831379744829e-05,
+      "loss": 2.1124,
+      "step": 17347
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.506631446757441e-05,
+      "loss": 2.1873,
+      "step": 17348
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5065797537191786e-05,
+      "loss": 2.0775,
+      "step": 17349
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.506528058859882e-05,
+      "loss": 1.8505,
+      "step": 17350
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.5064763621797367e-05,
+      "loss": 1.83,
+      "step": 17351
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5064246636789288e-05,
+      "loss": 1.8239,
+      "step": 17352
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.506372963357644e-05,
+      "loss": 1.7967,
+      "step": 17353
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.506321261216068e-05,
+      "loss": 2.0949,
+      "step": 17354
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5062695572543872e-05,
+      "loss": 1.8455,
+      "step": 17355
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5062178514727866e-05,
+      "loss": 1.8801,
+      "step": 17356
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5061661438714525e-05,
+      "loss": 1.9235,
+      "step": 17357
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5061144344505714e-05,
+      "loss": 2.0628,
+      "step": 17358
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5060627232103282e-05,
+      "loss": 2.1796,
+      "step": 17359
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5060110101509092e-05,
+      "loss": 1.9567,
+      "step": 17360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5059592952725002e-05,
+      "loss": 2.1549,
+      "step": 17361
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.5059075785752874e-05,
+      "loss": 1.8298,
+      "step": 17362
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5058558600594561e-05,
+      "loss": 1.6889,
+      "step": 17363
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.505804139725193e-05,
+      "loss": 1.8606,
+      "step": 17364
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5057524175726836e-05,
+      "loss": 1.8776,
+      "step": 17365
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5057006936021137e-05,
+      "loss": 1.9473,
+      "step": 17366
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5056489678136696e-05,
+      "loss": 1.8228,
+      "step": 17367
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5055972402075369e-05,
+      "loss": 1.7364,
+      "step": 17368
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.505545510783902e-05,
+      "loss": 1.7083,
+      "step": 17369
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5054937795429501e-05,
+      "loss": 1.9475,
+      "step": 17370
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5054420464848679e-05,
+      "loss": 1.9114,
+      "step": 17371
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.505390311609841e-05,
+      "loss": 1.8662,
+      "step": 17372
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5053385749180553e-05,
+      "loss": 1.9809,
+      "step": 17373
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5,
+      "learning_rate": 1.505286836409697e-05,
+      "loss": 1.6365,
+      "step": 17374
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5052350960849522e-05,
+      "loss": 1.7381,
+      "step": 17375
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5051833539440067e-05,
+      "loss": 1.8419,
+      "step": 17376
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5051316099870463e-05,
+      "loss": 1.816,
+      "step": 17377
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5050798642142572e-05,
+      "loss": 1.5565,
+      "step": 17378
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5050281166258258e-05,
+      "loss": 2.09,
+      "step": 17379
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.5049763672219375e-05,
+      "loss": 1.8455,
+      "step": 17380
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5049246160027786e-05,
+      "loss": 1.8582,
+      "step": 17381
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5048728629685352e-05,
+      "loss": 2.1693,
+      "step": 17382
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5048211081193932e-05,
+      "loss": 1.9138,
+      "step": 17383
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.5047693514555386e-05,
+      "loss": 1.8336,
+      "step": 17384
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5047175929771577e-05,
+      "loss": 1.9015,
+      "step": 17385
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.5046658326844364e-05,
+      "loss": 1.9119,
+      "step": 17386
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5046140705775605e-05,
+      "loss": 1.8374,
+      "step": 17387
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5045623066567166e-05,
+      "loss": 1.8987,
+      "step": 17388
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5045105409220905e-05,
+      "loss": 1.9502,
+      "step": 17389
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5044587733738685e-05,
+      "loss": 1.8767,
+      "step": 17390
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.5044070040122365e-05,
+      "loss": 1.6755,
+      "step": 17391
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5043552328373804e-05,
+      "loss": 2.1588,
+      "step": 17392
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5043034598494866e-05,
+      "loss": 2.4401,
+      "step": 17393
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.504251685048741e-05,
+      "loss": 1.8402,
+      "step": 17394
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.50419990843533e-05,
+      "loss": 2.2013,
+      "step": 17395
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5041481300094396e-05,
+      "loss": 1.6968,
+      "step": 17396
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5040963497712558e-05,
+      "loss": 1.7726,
+      "step": 17397
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5040445677209647e-05,
+      "loss": 1.6826,
+      "step": 17398
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.503992783858753e-05,
+      "loss": 2.0222,
+      "step": 17399
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5039409981848061e-05,
+      "loss": 1.9565,
+      "step": 17400
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5038892106993103e-05,
+      "loss": 1.7563,
+      "step": 17401
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.5038374214024525e-05,
+      "loss": 1.6289,
+      "step": 17402
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5037856302944178e-05,
+      "loss": 1.9601,
+      "step": 17403
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.5037338373753932e-05,
+      "loss": 1.54,
+      "step": 17404
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5036820426455645e-05,
+      "loss": 1.9148,
+      "step": 17405
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.503630246105118e-05,
+      "loss": 1.7135,
+      "step": 17406
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.50357844775424e-05,
+      "loss": 1.9477,
+      "step": 17407
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.5,
+      "learning_rate": 1.5035266475931162e-05,
+      "loss": 1.8409,
+      "step": 17408
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5034748456219336e-05,
+      "loss": 1.8522,
+      "step": 17409
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.5034230418408778e-05,
+      "loss": 1.9075,
+      "step": 17410
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.503371236250135e-05,
+      "loss": 2.0857,
+      "step": 17411
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5033194288498923e-05,
+      "loss": 1.4623,
+      "step": 17412
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.503267619640335e-05,
+      "loss": 2.1202,
+      "step": 17413
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5032158086216494e-05,
+      "loss": 2.098,
+      "step": 17414
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5031639957940221e-05,
+      "loss": 1.6019,
+      "step": 17415
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5031121811576392e-05,
+      "loss": 1.8697,
+      "step": 17416
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.503060364712687e-05,
+      "loss": 1.8763,
+      "step": 17417
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.5030085464593515e-05,
+      "loss": 2.3185,
+      "step": 17418
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.5029567263978195e-05,
+      "loss": 1.8066,
+      "step": 17419
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5029049045282773e-05,
+      "loss": 1.8926,
+      "step": 17420
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5028530808509105e-05,
+      "loss": 1.7937,
+      "step": 17421
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.502801255365906e-05,
+      "loss": 1.9267,
+      "step": 17422
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.5027494280734498e-05,
+      "loss": 2.297,
+      "step": 17423
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.502697598973728e-05,
+      "loss": 1.8482,
+      "step": 17424
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5026457680669278e-05,
+      "loss": 1.8968,
+      "step": 17425
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.5025939353532346e-05,
+      "loss": 2.0842,
+      "step": 17426
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.502542100832835e-05,
+      "loss": 2.1671,
+      "step": 17427
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5024902645059157e-05,
+      "loss": 1.5815,
+      "step": 17428
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.5024384263726626e-05,
+      "loss": 2.2202,
+      "step": 17429
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.502386586433262e-05,
+      "loss": 1.7666,
+      "step": 17430
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5023347446879007e-05,
+      "loss": 1.9526,
+      "step": 17431
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5022829011367645e-05,
+      "loss": 2.0016,
+      "step": 17432
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.5022310557800404e-05,
+      "loss": 2.0057,
+      "step": 17433
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5021792086179142e-05,
+      "loss": 2.0741,
+      "step": 17434
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5021273596505725e-05,
+      "loss": 1.7988,
+      "step": 17435
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.5020755088782018e-05,
+      "loss": 1.9868,
+      "step": 17436
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5020236563009882e-05,
+      "loss": 1.7654,
+      "step": 17437
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5019718019191185e-05,
+      "loss": 1.6994,
+      "step": 17438
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5019199457327788e-05,
+      "loss": 1.4468,
+      "step": 17439
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.5018680877421555e-05,
+      "loss": 2.0376,
+      "step": 17440
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5018162279474351e-05,
+      "loss": 2.0585,
+      "step": 17441
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5017643663488044e-05,
+      "loss": 1.8764,
+      "step": 17442
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.501712502946449e-05,
+      "loss": 1.9896,
+      "step": 17443
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.5016606377405563e-05,
+      "loss": 1.7487,
+      "step": 17444
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5016087707313116e-05,
+      "loss": 1.9025,
+      "step": 17445
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5015569019189027e-05,
+      "loss": 2.1394,
+      "step": 17446
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5015050313035149e-05,
+      "loss": 2.1887,
+      "step": 17447
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.5014531588853349e-05,
+      "loss": 2.195,
+      "step": 17448
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.5014012846645498e-05,
+      "loss": 2.2754,
+      "step": 17449
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5013494086413456e-05,
+      "loss": 2.0391,
+      "step": 17450
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5012975308159087e-05,
+      "loss": 1.6877,
+      "step": 17451
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5012456511884262e-05,
+      "loss": 1.8636,
+      "step": 17452
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.5011937697590837e-05,
+      "loss": 1.8008,
+      "step": 17453
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.501141886528068e-05,
+      "loss": 2.0121,
+      "step": 17454
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.501090001495566e-05,
+      "loss": 1.8609,
+      "step": 17455
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.5010381146617641e-05,
+      "loss": 1.9352,
+      "step": 17456
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5009862260268482e-05,
+      "loss": 1.8058,
+      "step": 17457
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5009343355910056e-05,
+      "loss": 1.9072,
+      "step": 17458
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5008824433544224e-05,
+      "loss": 1.7764,
+      "step": 17459
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5008305493172855e-05,
+      "loss": 1.984,
+      "step": 17460
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.5007786534797808e-05,
+      "loss": 1.9435,
+      "step": 17461
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5007267558420956e-05,
+      "loss": 1.9051,
+      "step": 17462
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.5006748564044162e-05,
+      "loss": 1.8227,
+      "step": 17463
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.5006229551669289e-05,
+      "loss": 1.961,
+      "step": 17464
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5005710521298204e-05,
+      "loss": 1.6291,
+      "step": 17465
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5005191472932776e-05,
+      "loss": 2.0227,
+      "step": 17466
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.5004672406574867e-05,
+      "loss": 1.998,
+      "step": 17467
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5004153322226342e-05,
+      "loss": 1.7673,
+      "step": 17468
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5003634219889077e-05,
+      "loss": 1.9539,
+      "step": 17469
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.5003115099564923e-05,
+      "loss": 1.7421,
+      "step": 17470
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.5002595961255754e-05,
+      "loss": 1.9262,
+      "step": 17471
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.5002076804963437e-05,
+      "loss": 1.6143,
+      "step": 17472
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.5001557630689836e-05,
+      "loss": 2.0532,
+      "step": 17473
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.500103843843682e-05,
+      "loss": 1.9716,
+      "step": 17474
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5000519228206255e-05,
+      "loss": 2.064,
+      "step": 17475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.7888,
+      "step": 17476
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4999480753819932e-05,
+      "loss": 1.8215,
+      "step": 17477
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4998961489667912e-05,
+      "loss": 1.8487,
+      "step": 17478
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.499844220754581e-05,
+      "loss": 1.8971,
+      "step": 17479
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4997922907455486e-05,
+      "loss": 1.6371,
+      "step": 17480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4997403589398813e-05,
+      "loss": 1.7295,
+      "step": 17481
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4996884253377656e-05,
+      "loss": 1.7638,
+      "step": 17482
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4996364899393882e-05,
+      "loss": 2.1117,
+      "step": 17483
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4995845527449357e-05,
+      "loss": 1.6157,
+      "step": 17484
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.499532613754595e-05,
+      "loss": 1.9239,
+      "step": 17485
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4994806729685527e-05,
+      "loss": 1.4781,
+      "step": 17486
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4994287303869954e-05,
+      "loss": 1.9192,
+      "step": 17487
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.49937678601011e-05,
+      "loss": 1.9745,
+      "step": 17488
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.499324839838083e-05,
+      "loss": 1.8304,
+      "step": 17489
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4992728918711014e-05,
+      "loss": 1.7093,
+      "step": 17490
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.499220942109352e-05,
+      "loss": 1.8598,
+      "step": 17491
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.499168990553021e-05,
+      "loss": 2.0785,
+      "step": 17492
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.499117037202296e-05,
+      "loss": 1.8029,
+      "step": 17493
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4990650820573629e-05,
+      "loss": 1.8856,
+      "step": 17494
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.499013125118409e-05,
+      "loss": 1.7195,
+      "step": 17495
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4989611663856209e-05,
+      "loss": 1.5822,
+      "step": 17496
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4989092058591856e-05,
+      "loss": 1.9221,
+      "step": 17497
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4988572435392895e-05,
+      "loss": 1.5698,
+      "step": 17498
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4988052794261197e-05,
+      "loss": 1.7316,
+      "step": 17499
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.4987533135198629e-05,
+      "loss": 1.7627,
+      "step": 17500
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4987013458207057e-05,
+      "loss": 1.8623,
+      "step": 17501
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.498649376328835e-05,
+      "loss": 1.5993,
+      "step": 17502
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4985974050444379e-05,
+      "loss": 1.7355,
+      "step": 17503
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.498545431967701e-05,
+      "loss": 1.8325,
+      "step": 17504
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4984934570988112e-05,
+      "loss": 2.0399,
+      "step": 17505
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4984414804379555e-05,
+      "loss": 1.9968,
+      "step": 17506
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4983895019853202e-05,
+      "loss": 1.9473,
+      "step": 17507
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4983375217410928e-05,
+      "loss": 2.0568,
+      "step": 17508
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4982855397054597e-05,
+      "loss": 1.4415,
+      "step": 17509
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4982335558786081e-05,
+      "loss": 1.9925,
+      "step": 17510
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4981815702607245e-05,
+      "loss": 1.8,
+      "step": 17511
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4981295828519961e-05,
+      "loss": 1.7548,
+      "step": 17512
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4980775936526095e-05,
+      "loss": 1.6941,
+      "step": 17513
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.498025602662752e-05,
+      "loss": 1.9051,
+      "step": 17514
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4979736098826099e-05,
+      "loss": 1.9421,
+      "step": 17515
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4979216153123708e-05,
+      "loss": 1.8045,
+      "step": 17516
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.4978696189522208e-05,
+      "loss": 1.9955,
+      "step": 17517
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4978176208023478e-05,
+      "loss": 1.9328,
+      "step": 17518
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4977656208629379e-05,
+      "loss": 1.7236,
+      "step": 17519
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4977136191341783e-05,
+      "loss": 1.7706,
+      "step": 17520
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.497661615616256e-05,
+      "loss": 1.5597,
+      "step": 17521
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4976096103093576e-05,
+      "loss": 1.9108,
+      "step": 17522
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.497557603213671e-05,
+      "loss": 1.9489,
+      "step": 17523
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4975055943293817e-05,
+      "loss": 2.2131,
+      "step": 17524
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4974535836566779e-05,
+      "loss": 1.6171,
+      "step": 17525
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.497401571195746e-05,
+      "loss": 1.76,
+      "step": 17526
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4973495569467733e-05,
+      "loss": 1.728,
+      "step": 17527
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4972975409099463e-05,
+      "loss": 1.8448,
+      "step": 17528
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4972455230854523e-05,
+      "loss": 1.8666,
+      "step": 17529
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4971935034734784e-05,
+      "loss": 1.7824,
+      "step": 17530
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4971414820742112e-05,
+      "loss": 1.7299,
+      "step": 17531
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.497089458887838e-05,
+      "loss": 2.3183,
+      "step": 17532
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4970374339145458e-05,
+      "loss": 1.731,
+      "step": 17533
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4969854071545219e-05,
+      "loss": 2.0675,
+      "step": 17534
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4969333786079526e-05,
+      "loss": 2.0053,
+      "step": 17535
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4968813482750254e-05,
+      "loss": 1.8042,
+      "step": 17536
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4968293161559273e-05,
+      "loss": 1.8097,
+      "step": 17537
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.496777282250845e-05,
+      "loss": 1.9413,
+      "step": 17538
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.496725246559966e-05,
+      "loss": 1.8749,
+      "step": 17539
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4966732090834778e-05,
+      "loss": 1.643,
+      "step": 17540
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.496621169821566e-05,
+      "loss": 2.5117,
+      "step": 17541
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.496569128774419e-05,
+      "loss": 1.8955,
+      "step": 17542
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.4965170859422235e-05,
+      "loss": 2.4452,
+      "step": 17543
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4964650413251662e-05,
+      "loss": 1.9675,
+      "step": 17544
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4964129949234347e-05,
+      "loss": 1.8788,
+      "step": 17545
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.496360946737216e-05,
+      "loss": 1.7089,
+      "step": 17546
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.496308896766697e-05,
+      "loss": 2.0024,
+      "step": 17547
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4962568450120644e-05,
+      "loss": 1.7582,
+      "step": 17548
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4962047914735063e-05,
+      "loss": 1.6348,
+      "step": 17549
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4961527361512094e-05,
+      "loss": 1.811,
+      "step": 17550
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.4961006790453605e-05,
+      "loss": 1.8251,
+      "step": 17551
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4960486201561468e-05,
+      "loss": 1.9171,
+      "step": 17552
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4959965594837563e-05,
+      "loss": 1.7699,
+      "step": 17553
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4959444970283749e-05,
+      "loss": 1.8465,
+      "step": 17554
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4958924327901904e-05,
+      "loss": 1.8871,
+      "step": 17555
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.49584036676939e-05,
+      "loss": 1.9874,
+      "step": 17556
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4957882989661607e-05,
+      "loss": 1.9817,
+      "step": 17557
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4957362293806896e-05,
+      "loss": 2.0657,
+      "step": 17558
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4956841580131642e-05,
+      "loss": 2.136,
+      "step": 17559
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4956320848637715e-05,
+      "loss": 1.7761,
+      "step": 17560
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4955800099326985e-05,
+      "loss": 2.1485,
+      "step": 17561
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4955279332201324e-05,
+      "loss": 1.6526,
+      "step": 17562
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4954758547262612e-05,
+      "loss": 2.0159,
+      "step": 17563
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4954237744512707e-05,
+      "loss": 1.6914,
+      "step": 17564
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4953716923953494e-05,
+      "loss": 1.9593,
+      "step": 17565
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4953196085586838e-05,
+      "loss": 1.8305,
+      "step": 17566
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4952675229414615e-05,
+      "loss": 1.9822,
+      "step": 17567
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4952154355438696e-05,
+      "loss": 2.1229,
+      "step": 17568
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4951633463660951e-05,
+      "loss": 2.013,
+      "step": 17569
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4951112554083258e-05,
+      "loss": 1.9789,
+      "step": 17570
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4950591626707483e-05,
+      "loss": 1.7438,
+      "step": 17571
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4950070681535502e-05,
+      "loss": 2.0593,
+      "step": 17572
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.494954971856919e-05,
+      "loss": 1.9342,
+      "step": 17573
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4949028737810416e-05,
+      "loss": 2.1171,
+      "step": 17574
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4948507739261056e-05,
+      "loss": 1.6535,
+      "step": 17575
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4947986722922976e-05,
+      "loss": 1.9206,
+      "step": 17576
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.494746568879806e-05,
+      "loss": 2.0456,
+      "step": 17577
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.494694463688817e-05,
+      "loss": 1.7617,
+      "step": 17578
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4946423567195185e-05,
+      "loss": 1.9906,
+      "step": 17579
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.494590247972098e-05,
+      "loss": 2.0552,
+      "step": 17580
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4945381374467422e-05,
+      "loss": 1.7905,
+      "step": 17581
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4944860251436391e-05,
+      "loss": 1.814,
+      "step": 17582
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4944339110629754e-05,
+      "loss": 1.7937,
+      "step": 17583
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4943817952049387e-05,
+      "loss": 1.8511,
+      "step": 17584
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4943296775697163e-05,
+      "loss": 1.7482,
+      "step": 17585
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4942775581574958e-05,
+      "loss": 1.6518,
+      "step": 17586
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4942254369684643e-05,
+      "loss": 1.8734,
+      "step": 17587
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4941733140028092e-05,
+      "loss": 2.24,
+      "step": 17588
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4941211892607181e-05,
+      "loss": 1.7607,
+      "step": 17589
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4940690627423782e-05,
+      "loss": 2.1195,
+      "step": 17590
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4940169344479767e-05,
+      "loss": 2.0559,
+      "step": 17591
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4939648043777009e-05,
+      "loss": 1.8512,
+      "step": 17592
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.4939126725317388e-05,
+      "loss": 1.6211,
+      "step": 17593
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4938605389102773e-05,
+      "loss": 1.8171,
+      "step": 17594
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4938084035135042e-05,
+      "loss": 1.8671,
+      "step": 17595
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4937562663416062e-05,
+      "loss": 1.8693,
+      "step": 17596
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4937041273947716e-05,
+      "loss": 1.8166,
+      "step": 17597
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4936519866731873e-05,
+      "loss": 2.3586,
+      "step": 17598
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4935998441770408e-05,
+      "loss": 1.8535,
+      "step": 17599
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.49354769990652e-05,
+      "loss": 1.9557,
+      "step": 17600
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4934955538618113e-05,
+      "loss": 1.6253,
+      "step": 17601
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.493443406043103e-05,
+      "loss": 1.6304,
+      "step": 17602
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4933912564505827e-05,
+      "loss": 1.911,
+      "step": 17603
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.493339105084437e-05,
+      "loss": 1.7751,
+      "step": 17604
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4932869519448543e-05,
+      "loss": 1.6798,
+      "step": 17605
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4932347970320214e-05,
+      "loss": 1.8225,
+      "step": 17606
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4931826403461262e-05,
+      "loss": 1.8698,
+      "step": 17607
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4931304818873557e-05,
+      "loss": 2.1821,
+      "step": 17608
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4930783216558982e-05,
+      "loss": 1.6842,
+      "step": 17609
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4930261596519406e-05,
+      "loss": 1.793,
+      "step": 17610
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4929739958756705e-05,
+      "loss": 1.7705,
+      "step": 17611
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.4929218303272757e-05,
+      "loss": 1.8878,
+      "step": 17612
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4928696630069431e-05,
+      "loss": 1.881,
+      "step": 17613
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.492817493914861e-05,
+      "loss": 1.7983,
+      "step": 17614
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4927653230512162e-05,
+      "loss": 1.8369,
+      "step": 17615
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4927131504161967e-05,
+      "loss": 2.0354,
+      "step": 17616
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4926609760099904e-05,
+      "loss": 1.9033,
+      "step": 17617
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4926087998327838e-05,
+      "loss": 1.7684,
+      "step": 17618
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4925566218847654e-05,
+      "loss": 1.6516,
+      "step": 17619
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4925044421661223e-05,
+      "loss": 2.0306,
+      "step": 17620
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4924522606770421e-05,
+      "loss": 1.9895,
+      "step": 17621
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4924000774177126e-05,
+      "loss": 1.8024,
+      "step": 17622
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4923478923883212e-05,
+      "loss": 1.9527,
+      "step": 17623
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.492295705589056e-05,
+      "loss": 2.0521,
+      "step": 17624
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4922435170201036e-05,
+      "loss": 1.506,
+      "step": 17625
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4921913266816523e-05,
+      "loss": 1.9142,
+      "step": 17626
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4921391345738898e-05,
+      "loss": 1.8514,
+      "step": 17627
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4920869406970034e-05,
+      "loss": 1.637,
+      "step": 17628
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4920347450511805e-05,
+      "loss": 1.6901,
+      "step": 17629
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4919825476366097e-05,
+      "loss": 1.7642,
+      "step": 17630
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4919303484534775e-05,
+      "loss": 1.8296,
+      "step": 17631
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4918781475019719e-05,
+      "loss": 1.8144,
+      "step": 17632
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.491825944782281e-05,
+      "loss": 2.4569,
+      "step": 17633
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4917737402945922e-05,
+      "loss": 1.8311,
+      "step": 17634
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.491721534039093e-05,
+      "loss": 2.0144,
+      "step": 17635
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.4916693260159712e-05,
+      "loss": 1.8415,
+      "step": 17636
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4916171162254145e-05,
+      "loss": 1.9831,
+      "step": 17637
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4915649046676104e-05,
+      "loss": 1.8811,
+      "step": 17638
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4915126913427464e-05,
+      "loss": 1.8191,
+      "step": 17639
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.491460476251011e-05,
+      "loss": 1.9202,
+      "step": 17640
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4914082593925914e-05,
+      "loss": 2.1067,
+      "step": 17641
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4913560407676752e-05,
+      "loss": 1.6349,
+      "step": 17642
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4913038203764502e-05,
+      "loss": 1.7345,
+      "step": 17643
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4912515982191043e-05,
+      "loss": 1.7968,
+      "step": 17644
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.4911993742958251e-05,
+      "loss": 2.0853,
+      "step": 17645
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4911471486068e-05,
+      "loss": 1.8112,
+      "step": 17646
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4910949211522174e-05,
+      "loss": 1.9636,
+      "step": 17647
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.4910426919322645e-05,
+      "loss": 1.8954,
+      "step": 17648
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4909904609471294e-05,
+      "loss": 1.7756,
+      "step": 17649
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4909382281969995e-05,
+      "loss": 1.7471,
+      "step": 17650
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.490885993682063e-05,
+      "loss": 1.8751,
+      "step": 17651
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4908337574025074e-05,
+      "loss": 1.5384,
+      "step": 17652
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4907815193585202e-05,
+      "loss": 1.8834,
+      "step": 17653
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4907292795502898e-05,
+      "loss": 1.5725,
+      "step": 17654
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4906770379780036e-05,
+      "loss": 1.8296,
+      "step": 17655
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4906247946418498e-05,
+      "loss": 2.0122,
+      "step": 17656
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4905725495420154e-05,
+      "loss": 1.8889,
+      "step": 17657
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.490520302678689e-05,
+      "loss": 1.936,
+      "step": 17658
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.490468054052058e-05,
+      "loss": 1.7717,
+      "step": 17659
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4904158036623101e-05,
+      "loss": 1.878,
+      "step": 17660
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4903635515096339e-05,
+      "loss": 2.0493,
+      "step": 17661
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4903112975942164e-05,
+      "loss": 1.8224,
+      "step": 17662
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.490259041916246e-05,
+      "loss": 1.7301,
+      "step": 17663
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.49020678447591e-05,
+      "loss": 1.8556,
+      "step": 17664
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4901545252733967e-05,
+      "loss": 1.7381,
+      "step": 17665
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4901022643088939e-05,
+      "loss": 1.7975,
+      "step": 17666
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.490050001582589e-05,
+      "loss": 1.8044,
+      "step": 17667
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4899977370946705e-05,
+      "loss": 1.852,
+      "step": 17668
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4899454708453257e-05,
+      "loss": 1.5939,
+      "step": 17669
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4898932028347433e-05,
+      "loss": 2.0155,
+      "step": 17670
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4898409330631102e-05,
+      "loss": 2.1257,
+      "step": 17671
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4897886615306151e-05,
+      "loss": 1.796,
+      "step": 17672
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4897363882374456e-05,
+      "loss": 1.7902,
+      "step": 17673
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4896841131837896e-05,
+      "loss": 1.9638,
+      "step": 17674
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.489631836369835e-05,
+      "loss": 1.9167,
+      "step": 17675
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4895795577957695e-05,
+      "loss": 1.8266,
+      "step": 17676
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4895272774617815e-05,
+      "loss": 2.2072,
+      "step": 17677
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4894749953680588e-05,
+      "loss": 1.5495,
+      "step": 17678
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4894227115147891e-05,
+      "loss": 1.8151,
+      "step": 17679
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4893704259021603e-05,
+      "loss": 2.1068,
+      "step": 17680
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.489318138530361e-05,
+      "loss": 1.9308,
+      "step": 17681
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4892658493995784e-05,
+      "loss": 1.7838,
+      "step": 17682
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4892135585100007e-05,
+      "loss": 1.9675,
+      "step": 17683
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4891612658618164e-05,
+      "loss": 2.193,
+      "step": 17684
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4891089714552124e-05,
+      "loss": 2.0501,
+      "step": 17685
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4890566752903776e-05,
+      "loss": 2.1358,
+      "step": 17686
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4890043773675e-05,
+      "loss": 1.6929,
+      "step": 17687
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4889520776867672e-05,
+      "loss": 1.7315,
+      "step": 17688
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.488899776248367e-05,
+      "loss": 1.9844,
+      "step": 17689
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4888474730524877e-05,
+      "loss": 1.737,
+      "step": 17690
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.488795168099318e-05,
+      "loss": 2.9684,
+      "step": 17691
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4887428613890446e-05,
+      "loss": 1.5957,
+      "step": 17692
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4886905529218566e-05,
+      "loss": 1.6857,
+      "step": 17693
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4886382426979413e-05,
+      "loss": 1.7544,
+      "step": 17694
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4885859307174872e-05,
+      "loss": 2.2013,
+      "step": 17695
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4885336169806826e-05,
+      "loss": 1.921,
+      "step": 17696
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.4884813014877147e-05,
+      "loss": 1.917,
+      "step": 17697
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4884289842387723e-05,
+      "loss": 1.8408,
+      "step": 17698
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4883766652340427e-05,
+      "loss": 1.7798,
+      "step": 17699
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.488324344473715e-05,
+      "loss": 2.2585,
+      "step": 17700
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.488272021957977e-05,
+      "loss": 1.8982,
+      "step": 17701
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.488219697687016e-05,
+      "loss": 1.8239,
+      "step": 17702
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.488167371661021e-05,
+      "loss": 1.9528,
+      "step": 17703
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.4881150438801797e-05,
+      "loss": 2.3484,
+      "step": 17704
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4880627143446802e-05,
+      "loss": 1.9505,
+      "step": 17705
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4880103830547104e-05,
+      "loss": 1.9728,
+      "step": 17706
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4879580500104589e-05,
+      "loss": 1.6953,
+      "step": 17707
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4879057152121138e-05,
+      "loss": 1.8543,
+      "step": 17708
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.4878533786598626e-05,
+      "loss": 2.0368,
+      "step": 17709
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4878010403538944e-05,
+      "loss": 1.8,
+      "step": 17710
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4877487002943965e-05,
+      "loss": 1.8869,
+      "step": 17711
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4876963584815574e-05,
+      "loss": 1.8067,
+      "step": 17712
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4876440149155651e-05,
+      "loss": 2.5147,
+      "step": 17713
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4875916695966081e-05,
+      "loss": 1.6604,
+      "step": 17714
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4875393225248745e-05,
+      "loss": 1.8382,
+      "step": 17715
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.487486973700552e-05,
+      "loss": 1.9106,
+      "step": 17716
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4874346231238292e-05,
+      "loss": 1.7925,
+      "step": 17717
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4873822707948945e-05,
+      "loss": 1.7326,
+      "step": 17718
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4873299167139354e-05,
+      "loss": 1.8103,
+      "step": 17719
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4872775608811406e-05,
+      "loss": 1.7596,
+      "step": 17720
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4872252032966985e-05,
+      "loss": 1.8692,
+      "step": 17721
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4871728439607967e-05,
+      "loss": 2.0932,
+      "step": 17722
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.487120482873624e-05,
+      "loss": 1.8032,
+      "step": 17723
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4870681200353678e-05,
+      "loss": 1.9958,
+      "step": 17724
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4870157554462175e-05,
+      "loss": 1.891,
+      "step": 17725
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4869633891063605e-05,
+      "loss": 1.6044,
+      "step": 17726
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4869110210159854e-05,
+      "loss": 1.8545,
+      "step": 17727
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.4868586511752802e-05,
+      "loss": 1.8105,
+      "step": 17728
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4868062795844334e-05,
+      "loss": 1.7968,
+      "step": 17729
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.486753906243633e-05,
+      "loss": 1.9049,
+      "step": 17730
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4867015311530677e-05,
+      "loss": 2.0711,
+      "step": 17731
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4866491543129252e-05,
+      "loss": 1.9301,
+      "step": 17732
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.486596775723394e-05,
+      "loss": 1.8347,
+      "step": 17733
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4865443953846629e-05,
+      "loss": 1.855,
+      "step": 17734
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4864920132969194e-05,
+      "loss": 2.1195,
+      "step": 17735
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4864396294603523e-05,
+      "loss": 1.8622,
+      "step": 17736
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4863872438751496e-05,
+      "loss": 1.7517,
+      "step": 17737
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4863348565415e-05,
+      "loss": 2.1275,
+      "step": 17738
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4862824674595915e-05,
+      "loss": 2.1324,
+      "step": 17739
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4862300766296125e-05,
+      "loss": 1.9599,
+      "step": 17740
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4861776840517514e-05,
+      "loss": 1.7489,
+      "step": 17741
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4861252897261967e-05,
+      "loss": 1.9115,
+      "step": 17742
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4860728936531363e-05,
+      "loss": 1.5971,
+      "step": 17743
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.486020495832759e-05,
+      "loss": 1.918,
+      "step": 17744
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.485968096265253e-05,
+      "loss": 2.1614,
+      "step": 17745
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4859156949508062e-05,
+      "loss": 1.7614,
+      "step": 17746
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4858632918896076e-05,
+      "loss": 2.0631,
+      "step": 17747
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4858108870818455e-05,
+      "loss": 1.9047,
+      "step": 17748
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.485758480527708e-05,
+      "loss": 2.0468,
+      "step": 17749
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4857060722273836e-05,
+      "loss": 1.8038,
+      "step": 17750
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4856536621810611e-05,
+      "loss": 2.1279,
+      "step": 17751
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4856012503889283e-05,
+      "loss": 1.8947,
+      "step": 17752
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4855488368511737e-05,
+      "loss": 1.5993,
+      "step": 17753
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.485496421567986e-05,
+      "loss": 1.8412,
+      "step": 17754
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4854440045395536e-05,
+      "loss": 1.9711,
+      "step": 17755
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4853915857660645e-05,
+      "loss": 1.8776,
+      "step": 17756
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4853391652477078e-05,
+      "loss": 1.9393,
+      "step": 17757
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.4852867429846716e-05,
+      "loss": 3.0396,
+      "step": 17758
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4852343189771441e-05,
+      "loss": 1.8773,
+      "step": 17759
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4851818932253137e-05,
+      "loss": 1.8804,
+      "step": 17760
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4851294657293696e-05,
+      "loss": 2.1516,
+      "step": 17761
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4850770364894995e-05,
+      "loss": 1.9793,
+      "step": 17762
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4850246055058923e-05,
+      "loss": 2.0563,
+      "step": 17763
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4849721727787365e-05,
+      "loss": 1.821,
+      "step": 17764
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4849197383082204e-05,
+      "loss": 1.9053,
+      "step": 17765
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4848673020945326e-05,
+      "loss": 2.0106,
+      "step": 17766
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.484814864137861e-05,
+      "loss": 1.6743,
+      "step": 17767
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4847624244383951e-05,
+      "loss": 1.684,
+      "step": 17768
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4847099829963227e-05,
+      "loss": 1.8868,
+      "step": 17769
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4846575398118326e-05,
+      "loss": 1.9048,
+      "step": 17770
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4846050948851134e-05,
+      "loss": 1.75,
+      "step": 17771
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4845526482163536e-05,
+      "loss": 2.4147,
+      "step": 17772
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4845001998057411e-05,
+      "loss": 1.6332,
+      "step": 17773
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4844477496534653e-05,
+      "loss": 1.9211,
+      "step": 17774
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4843952977597144e-05,
+      "loss": 1.6452,
+      "step": 17775
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4843428441246768e-05,
+      "loss": 1.9111,
+      "step": 17776
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4842903887485416e-05,
+      "loss": 1.7984,
+      "step": 17777
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4842379316314967e-05,
+      "loss": 1.6196,
+      "step": 17778
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.484185472773731e-05,
+      "loss": 2.0718,
+      "step": 17779
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4841330121754332e-05,
+      "loss": 1.6963,
+      "step": 17780
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4840805498367912e-05,
+      "loss": 1.8351,
+      "step": 17781
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4840280857579947e-05,
+      "loss": 2.1958,
+      "step": 17782
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.4839756199392312e-05,
+      "loss": 1.5638,
+      "step": 17783
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.48392315238069e-05,
+      "loss": 1.6109,
+      "step": 17784
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4838706830825599e-05,
+      "loss": 2.3042,
+      "step": 17785
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4838182120450286e-05,
+      "loss": 1.8115,
+      "step": 17786
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4837657392682854e-05,
+      "loss": 1.8728,
+      "step": 17787
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.483713264752519e-05,
+      "loss": 1.8281,
+      "step": 17788
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4836607884979175e-05,
+      "loss": 2.0572,
+      "step": 17789
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4836083105046699e-05,
+      "loss": 1.9576,
+      "step": 17790
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4835558307729647e-05,
+      "loss": 1.725,
+      "step": 17791
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4835033493029911e-05,
+      "loss": 1.692,
+      "step": 17792
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4834508660949369e-05,
+      "loss": 1.7201,
+      "step": 17793
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4833983811489914e-05,
+      "loss": 1.9523,
+      "step": 17794
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.483345894465343e-05,
+      "loss": 1.4558,
+      "step": 17795
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4832934060441804e-05,
+      "loss": 1.9328,
+      "step": 17796
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.4832409158856921e-05,
+      "loss": 1.8996,
+      "step": 17797
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4831884239900672e-05,
+      "loss": 1.7145,
+      "step": 17798
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.483135930357494e-05,
+      "loss": 1.9913,
+      "step": 17799
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4830834349881613e-05,
+      "loss": 2.3733,
+      "step": 17800
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4830309378822583e-05,
+      "loss": 1.693,
+      "step": 17801
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4829784390399731e-05,
+      "loss": 1.6527,
+      "step": 17802
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4829259384614947e-05,
+      "loss": 1.7499,
+      "step": 17803
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4828734361470116e-05,
+      "loss": 2.041,
+      "step": 17804
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4828209320967131e-05,
+      "loss": 1.7656,
+      "step": 17805
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4827684263107871e-05,
+      "loss": 1.736,
+      "step": 17806
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4827159187894227e-05,
+      "loss": 1.9389,
+      "step": 17807
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4826634095328089e-05,
+      "loss": 1.7142,
+      "step": 17808
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4826108985411345e-05,
+      "loss": 1.8868,
+      "step": 17809
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.482558385814588e-05,
+      "loss": 1.8034,
+      "step": 17810
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.482505871353358e-05,
+      "loss": 1.8538,
+      "step": 17811
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4824533551576336e-05,
+      "loss": 2.1526,
+      "step": 17812
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4824008372276035e-05,
+      "loss": 2.0001,
+      "step": 17813
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4823483175634565e-05,
+      "loss": 1.9471,
+      "step": 17814
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4822957961653814e-05,
+      "loss": 1.9122,
+      "step": 17815
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.482243273033567e-05,
+      "loss": 1.7762,
+      "step": 17816
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4821907481682018e-05,
+      "loss": 1.6844,
+      "step": 17817
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4821382215694752e-05,
+      "loss": 1.8573,
+      "step": 17818
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4820856932375756e-05,
+      "loss": 1.8859,
+      "step": 17819
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.482033163172692e-05,
+      "loss": 1.6669,
+      "step": 17820
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.481980631375013e-05,
+      "loss": 1.8861,
+      "step": 17821
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4819280978447279e-05,
+      "loss": 1.94,
+      "step": 17822
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4818755625820249e-05,
+      "loss": 1.7062,
+      "step": 17823
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4818230255870935e-05,
+      "loss": 1.9162,
+      "step": 17824
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4817704868601221e-05,
+      "loss": 2.114,
+      "step": 17825
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4817179464012999e-05,
+      "loss": 1.8143,
+      "step": 17826
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4816654042108157e-05,
+      "loss": 1.8567,
+      "step": 17827
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4816128602888577e-05,
+      "loss": 1.9616,
+      "step": 17828
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.481560314635616e-05,
+      "loss": 1.8171,
+      "step": 17829
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.4815077672512788e-05,
+      "loss": 3.2247,
+      "step": 17830
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4814552181360346e-05,
+      "loss": 1.9455,
+      "step": 17831
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4814026672900732e-05,
+      "loss": 1.9505,
+      "step": 17832
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4813501147135829e-05,
+      "loss": 1.9333,
+      "step": 17833
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4812975604067528e-05,
+      "loss": 1.6462,
+      "step": 17834
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4812450043697717e-05,
+      "loss": 1.7685,
+      "step": 17835
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4811924466028287e-05,
+      "loss": 2.4085,
+      "step": 17836
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4811398871061125e-05,
+      "loss": 1.9535,
+      "step": 17837
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4810873258798124e-05,
+      "loss": 1.6077,
+      "step": 17838
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.481034762924117e-05,
+      "loss": 1.9272,
+      "step": 17839
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4809821982392155e-05,
+      "loss": 2.0426,
+      "step": 17840
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4809296318252968e-05,
+      "loss": 1.7818,
+      "step": 17841
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4808770636825498e-05,
+      "loss": 1.8496,
+      "step": 17842
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4808244938111634e-05,
+      "loss": 1.7116,
+      "step": 17843
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4807719222113265e-05,
+      "loss": 1.7039,
+      "step": 17844
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4807193488832284e-05,
+      "loss": 2.0695,
+      "step": 17845
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.480666773827058e-05,
+      "loss": 2.0513,
+      "step": 17846
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4806141970430041e-05,
+      "loss": 1.7778,
+      "step": 17847
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4805616185312558e-05,
+      "loss": 1.8206,
+      "step": 17848
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4805090382920022e-05,
+      "loss": 1.9934,
+      "step": 17849
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4804564563254325e-05,
+      "loss": 1.708,
+      "step": 17850
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4804038726317349e-05,
+      "loss": 1.749,
+      "step": 17851
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4803512872110995e-05,
+      "loss": 1.547,
+      "step": 17852
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4802987000637143e-05,
+      "loss": 1.8173,
+      "step": 17853
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4802461111897692e-05,
+      "loss": 1.842,
+      "step": 17854
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4801935205894531e-05,
+      "loss": 1.7993,
+      "step": 17855
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4801409282629546e-05,
+      "loss": 1.8021,
+      "step": 17856
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.480088334210463e-05,
+      "loss": 1.9778,
+      "step": 17857
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4800357384321673e-05,
+      "loss": 1.9001,
+      "step": 17858
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.479983140928257e-05,
+      "loss": 1.8831,
+      "step": 17859
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4799305416989205e-05,
+      "loss": 1.6924,
+      "step": 17860
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4798779407443472e-05,
+      "loss": 1.5892,
+      "step": 17861
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4798253380647265e-05,
+      "loss": 1.8041,
+      "step": 17862
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4797727336602468e-05,
+      "loss": 1.9577,
+      "step": 17863
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4797201275310977e-05,
+      "loss": 2.0075,
+      "step": 17864
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4796675196774681e-05,
+      "loss": 1.7643,
+      "step": 17865
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4796149100995477e-05,
+      "loss": 1.7995,
+      "step": 17866
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.4795622987975247e-05,
+      "loss": 1.5536,
+      "step": 17867
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4795096857715887e-05,
+      "loss": 1.8191,
+      "step": 17868
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.479457071021929e-05,
+      "loss": 2.1231,
+      "step": 17869
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.479404454548734e-05,
+      "loss": 1.9768,
+      "step": 17870
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4793518363521937e-05,
+      "loss": 1.7986,
+      "step": 17871
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4792992164324971e-05,
+      "loss": 1.989,
+      "step": 17872
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4792465947898329e-05,
+      "loss": 1.7246,
+      "step": 17873
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4791939714243904e-05,
+      "loss": 1.9923,
+      "step": 17874
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4791413463363591e-05,
+      "loss": 1.7854,
+      "step": 17875
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.4790887195259282e-05,
+      "loss": 1.7548,
+      "step": 17876
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4790360909932864e-05,
+      "loss": 1.5907,
+      "step": 17877
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.478983460738623e-05,
+      "loss": 1.7574,
+      "step": 17878
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4789308287621275e-05,
+      "loss": 1.8409,
+      "step": 17879
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.478878195063989e-05,
+      "loss": 1.8474,
+      "step": 17880
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4788255596443968e-05,
+      "loss": 1.8953,
+      "step": 17881
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4787729225035397e-05,
+      "loss": 1.6055,
+      "step": 17882
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4787202836416072e-05,
+      "loss": 1.9022,
+      "step": 17883
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4786676430587884e-05,
+      "loss": 2.0011,
+      "step": 17884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4786150007552726e-05,
+      "loss": 1.9143,
+      "step": 17885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4785623567312492e-05,
+      "loss": 2.1553,
+      "step": 17886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4785097109869072e-05,
+      "loss": 2.0159,
+      "step": 17887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4784570635224363e-05,
+      "loss": 1.8376,
+      "step": 17888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4784044143380252e-05,
+      "loss": 1.9381,
+      "step": 17889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4783517634338634e-05,
+      "loss": 2.0778,
+      "step": 17890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4782991108101397e-05,
+      "loss": 1.6542,
+      "step": 17891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4782464564670443e-05,
+      "loss": 1.9525,
+      "step": 17892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.4781938004047658e-05,
+      "loss": 1.806,
+      "step": 17893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4781411426234937e-05,
+      "loss": 1.9334,
+      "step": 17894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4780884831234174e-05,
+      "loss": 2.0244,
+      "step": 17895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4780358219047259e-05,
+      "loss": 2.0296,
+      "step": 17896
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4779831589676088e-05,
+      "loss": 1.8341,
+      "step": 17897
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.477930494312255e-05,
+      "loss": 1.6891,
+      "step": 17898
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4778778279388545e-05,
+      "loss": 1.6509,
+      "step": 17899
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4778251598475958e-05,
+      "loss": 1.7647,
+      "step": 17900
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4777724900386692e-05,
+      "loss": 2.0148,
+      "step": 17901
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.477719818512263e-05,
+      "loss": 2.5455,
+      "step": 17902
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4776671452685674e-05,
+      "loss": 1.7205,
+      "step": 17903
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.477614470307771e-05,
+      "loss": 1.7596,
+      "step": 17904
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4775617936300638e-05,
+      "loss": 2.0982,
+      "step": 17905
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4775091152356349e-05,
+      "loss": 1.8167,
+      "step": 17906
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4774564351246734e-05,
+      "loss": 2.0167,
+      "step": 17907
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4774037532973694e-05,
+      "loss": 1.8716,
+      "step": 17908
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4773510697539112e-05,
+      "loss": 1.9018,
+      "step": 17909
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4772983844944893e-05,
+      "loss": 1.6268,
+      "step": 17910
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4772456975192924e-05,
+      "loss": 1.9982,
+      "step": 17911
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4771930088285098e-05,
+      "loss": 1.9458,
+      "step": 17912
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4771403184223319e-05,
+      "loss": 1.7396,
+      "step": 17913
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4770876263009468e-05,
+      "loss": 2.0149,
+      "step": 17914
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4770349324645446e-05,
+      "loss": 1.6891,
+      "step": 17915
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4769822369133147e-05,
+      "loss": 2.3369,
+      "step": 17916
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4769295396474466e-05,
+      "loss": 1.9286,
+      "step": 17917
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4768768406671295e-05,
+      "loss": 2.127,
+      "step": 17918
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.476824139972553e-05,
+      "loss": 1.7939,
+      "step": 17919
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4767714375639064e-05,
+      "loss": 1.9403,
+      "step": 17920
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4767187334413789e-05,
+      "loss": 1.7063,
+      "step": 17921
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4766660276051608e-05,
+      "loss": 1.9305,
+      "step": 17922
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.4766133200554406e-05,
+      "loss": 1.7707,
+      "step": 17923
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4765606107924086e-05,
+      "loss": 1.7439,
+      "step": 17924
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.476507899816254e-05,
+      "loss": 1.75,
+      "step": 17925
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.476455187127166e-05,
+      "loss": 1.9738,
+      "step": 17926
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4764024727253343e-05,
+      "loss": 2.1203,
+      "step": 17927
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4763497566109482e-05,
+      "loss": 2.046,
+      "step": 17928
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4762970387841976e-05,
+      "loss": 1.7731,
+      "step": 17929
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4762443192452715e-05,
+      "loss": 1.9205,
+      "step": 17930
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.47619159799436e-05,
+      "loss": 1.6766,
+      "step": 17931
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4761388750316523e-05,
+      "loss": 1.6945,
+      "step": 17932
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.476086150357338e-05,
+      "loss": 2.0131,
+      "step": 17933
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4760334239716065e-05,
+      "loss": 1.825,
+      "step": 17934
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4759806958746471e-05,
+      "loss": 2.0528,
+      "step": 17935
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4759279660666504e-05,
+      "loss": 2.3355,
+      "step": 17936
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4758752345478045e-05,
+      "loss": 1.9702,
+      "step": 17937
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4758225013182998e-05,
+      "loss": 2.2553,
+      "step": 17938
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4757697663783262e-05,
+      "loss": 1.7264,
+      "step": 17939
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4757170297280725e-05,
+      "loss": 2.0988,
+      "step": 17940
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4756642913677283e-05,
+      "loss": 1.7624,
+      "step": 17941
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4756115512974837e-05,
+      "loss": 1.7469,
+      "step": 17942
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4755588095175284e-05,
+      "loss": 2.2001,
+      "step": 17943
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4755060660280511e-05,
+      "loss": 1.9182,
+      "step": 17944
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4754533208292423e-05,
+      "loss": 1.8884,
+      "step": 17945
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4754005739212914e-05,
+      "loss": 2.1197,
+      "step": 17946
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.4753478253043873e-05,
+      "loss": 2.0297,
+      "step": 17947
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4752950749787204e-05,
+      "loss": 1.7524,
+      "step": 17948
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.47524232294448e-05,
+      "loss": 1.9979,
+      "step": 17949
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4751895692018564e-05,
+      "loss": 1.9528,
+      "step": 17950
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.4751368137510383e-05,
+      "loss": 3.099,
+      "step": 17951
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4750840565922155e-05,
+      "loss": 1.6876,
+      "step": 17952
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4750312977255785e-05,
+      "loss": 2.0261,
+      "step": 17953
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4749785371513156e-05,
+      "loss": 2.1696,
+      "step": 17954
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4749257748696178e-05,
+      "loss": 2.1915,
+      "step": 17955
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4748730108806737e-05,
+      "loss": 1.7305,
+      "step": 17956
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4748202451846738e-05,
+      "loss": 1.7838,
+      "step": 17957
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.474767477781807e-05,
+      "loss": 1.9818,
+      "step": 17958
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4747147086722635e-05,
+      "loss": 1.8608,
+      "step": 17959
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4746619378562332e-05,
+      "loss": 2.0339,
+      "step": 17960
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4746091653339051e-05,
+      "loss": 1.8689,
+      "step": 17961
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4745563911054696e-05,
+      "loss": 1.8098,
+      "step": 17962
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.474503615171116e-05,
+      "loss": 1.9716,
+      "step": 17963
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4744508375310342e-05,
+      "loss": 2.0147,
+      "step": 17964
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4743980581854138e-05,
+      "loss": 2.0254,
+      "step": 17965
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4743452771344444e-05,
+      "loss": 1.9966,
+      "step": 17966
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4742924943783161e-05,
+      "loss": 1.9888,
+      "step": 17967
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4742397099172183e-05,
+      "loss": 1.7365,
+      "step": 17968
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4741869237513409e-05,
+      "loss": 1.5664,
+      "step": 17969
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4741341358808738e-05,
+      "loss": 1.8314,
+      "step": 17970
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4740813463060066e-05,
+      "loss": 1.9685,
+      "step": 17971
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4740285550269292e-05,
+      "loss": 2.3877,
+      "step": 17972
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4739757620438308e-05,
+      "loss": 1.7278,
+      "step": 17973
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.473922967356902e-05,
+      "loss": 1.6255,
+      "step": 17974
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4738701709663318e-05,
+      "loss": 1.9072,
+      "step": 17975
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4738173728723109e-05,
+      "loss": 1.865,
+      "step": 17976
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4737645730750285e-05,
+      "loss": 1.6886,
+      "step": 17977
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4737117715746743e-05,
+      "loss": 1.8212,
+      "step": 17978
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4736589683714385e-05,
+      "loss": 1.5673,
+      "step": 17979
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4736061634655107e-05,
+      "loss": 1.8858,
+      "step": 17980
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4735533568570808e-05,
+      "loss": 1.8484,
+      "step": 17981
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.4735005485463384e-05,
+      "loss": 1.6411,
+      "step": 17982
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4734477385334735e-05,
+      "loss": 1.843,
+      "step": 17983
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.473394926818676e-05,
+      "loss": 2.0268,
+      "step": 17984
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4733421134021357e-05,
+      "loss": 2.0738,
+      "step": 17985
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4732892982840425e-05,
+      "loss": 1.749,
+      "step": 17986
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4732364814645861e-05,
+      "loss": 1.9027,
+      "step": 17987
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.4731836629439567e-05,
+      "loss": 2.0745,
+      "step": 17988
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4731308427223435e-05,
+      "loss": 1.579,
+      "step": 17989
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4730780207999374e-05,
+      "loss": 1.8891,
+      "step": 17990
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4730251971769272e-05,
+      "loss": 1.7872,
+      "step": 17991
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4729723718535034e-05,
+      "loss": 1.8256,
+      "step": 17992
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.472919544829856e-05,
+      "loss": 1.6674,
+      "step": 17993
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4728667161061743e-05,
+      "loss": 2.1509,
+      "step": 17994
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4728138856826489e-05,
+      "loss": 1.9374,
+      "step": 17995
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4727610535594693e-05,
+      "loss": 1.9041,
+      "step": 17996
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4727082197368256e-05,
+      "loss": 1.9846,
+      "step": 17997
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4726553842149074e-05,
+      "loss": 2.1839,
+      "step": 17998
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4726025469939049e-05,
+      "loss": 2.0728,
+      "step": 17999
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4725497080740083e-05,
+      "loss": 1.8425,
+      "step": 18000
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4724968674554072e-05,
+      "loss": 1.8454,
+      "step": 18001
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4724440251382914e-05,
+      "loss": 1.8652,
+      "step": 18002
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.472391181122851e-05,
+      "loss": 2.0381,
+      "step": 18003
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4723383354092764e-05,
+      "loss": 1.878,
+      "step": 18004
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4722854879977569e-05,
+      "loss": 2.1232,
+      "step": 18005
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4722326388884826e-05,
+      "loss": 2.5012,
+      "step": 18006
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4721797880816442e-05,
+      "loss": 1.9386,
+      "step": 18007
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4721269355774307e-05,
+      "loss": 1.8549,
+      "step": 18008
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4720740813760326e-05,
+      "loss": 1.779,
+      "step": 18009
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.47202122547764e-05,
+      "loss": 2.1046,
+      "step": 18010
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4719683678824426e-05,
+      "loss": 1.9656,
+      "step": 18011
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4719155085906303e-05,
+      "loss": 1.8666,
+      "step": 18012
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4718626476023939e-05,
+      "loss": 1.6371,
+      "step": 18013
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4718097849179223e-05,
+      "loss": 2.1673,
+      "step": 18014
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4717569205374063e-05,
+      "loss": 1.9732,
+      "step": 18015
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4717040544610356e-05,
+      "loss": 1.8231,
+      "step": 18016
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4716511866890007e-05,
+      "loss": 1.8573,
+      "step": 18017
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4715983172214911e-05,
+      "loss": 1.7226,
+      "step": 18018
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4715454460586971e-05,
+      "loss": 2.3429,
+      "step": 18019
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.471492573200809e-05,
+      "loss": 1.8959,
+      "step": 18020
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4714396986480161e-05,
+      "loss": 1.8148,
+      "step": 18021
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4713868224005094e-05,
+      "loss": 2.0918,
+      "step": 18022
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4713339444584784e-05,
+      "loss": 1.6885,
+      "step": 18023
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4712810648221135e-05,
+      "loss": 1.8728,
+      "step": 18024
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4712281834916044e-05,
+      "loss": 1.7984,
+      "step": 18025
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4711753004671414e-05,
+      "loss": 1.6264,
+      "step": 18026
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.471122415748915e-05,
+      "loss": 1.8237,
+      "step": 18027
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4710695293371145e-05,
+      "loss": 1.9618,
+      "step": 18028
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4710166412319307e-05,
+      "loss": 1.7339,
+      "step": 18029
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4709637514335534e-05,
+      "loss": 2.2467,
+      "step": 18030
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4709108599421727e-05,
+      "loss": 1.9288,
+      "step": 18031
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4708579667579789e-05,
+      "loss": 1.8093,
+      "step": 18032
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.470805071881162e-05,
+      "loss": 1.9337,
+      "step": 18033
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4707521753119125e-05,
+      "loss": 1.9095,
+      "step": 18034
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4706992770504198e-05,
+      "loss": 1.992,
+      "step": 18035
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.470646377096875e-05,
+      "loss": 2.1922,
+      "step": 18036
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4705934754514676e-05,
+      "loss": 1.8762,
+      "step": 18037
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4705405721143877e-05,
+      "loss": 1.9212,
+      "step": 18038
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.470487667085826e-05,
+      "loss": 1.9451,
+      "step": 18039
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4704347603659722e-05,
+      "loss": 1.9342,
+      "step": 18040
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4703818519550168e-05,
+      "loss": 1.7949,
+      "step": 18041
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4703289418531499e-05,
+      "loss": 1.7984,
+      "step": 18042
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4702760300605616e-05,
+      "loss": 1.7304,
+      "step": 18043
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4702231165774423e-05,
+      "loss": 1.711,
+      "step": 18044
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.375,
+      "learning_rate": 1.470170201403982e-05,
+      "loss": 1.9686,
+      "step": 18045
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.470117284540371e-05,
+      "loss": 1.7039,
+      "step": 18046
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4700643659867997e-05,
+      "loss": 1.7749,
+      "step": 18047
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4700114457434582e-05,
+      "loss": 1.8819,
+      "step": 18048
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4699585238105366e-05,
+      "loss": 1.9061,
+      "step": 18049
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4699056001882252e-05,
+      "loss": 1.9263,
+      "step": 18050
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4698526748767143e-05,
+      "loss": 1.7865,
+      "step": 18051
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.469799747876194e-05,
+      "loss": 2.109,
+      "step": 18052
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4697468191868552e-05,
+      "loss": 1.9562,
+      "step": 18053
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4696938888088872e-05,
+      "loss": 1.9014,
+      "step": 18054
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.469640956742481e-05,
+      "loss": 2.027,
+      "step": 18055
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4695880229878265e-05,
+      "loss": 1.7035,
+      "step": 18056
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4695350875451145e-05,
+      "loss": 1.7756,
+      "step": 18057
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4694821504145344e-05,
+      "loss": 1.5804,
+      "step": 18058
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.469429211596277e-05,
+      "loss": 1.8156,
+      "step": 18059
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4693762710905328e-05,
+      "loss": 2.0546,
+      "step": 18060
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4693233288974918e-05,
+      "loss": 2.126,
+      "step": 18061
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4692703850173446e-05,
+      "loss": 1.8502,
+      "step": 18062
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4692174394502812e-05,
+      "loss": 1.7756,
+      "step": 18063
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4691644921964923e-05,
+      "loss": 1.7654,
+      "step": 18064
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4691115432561678e-05,
+      "loss": 1.8034,
+      "step": 18065
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.469058592629498e-05,
+      "loss": 1.8081,
+      "step": 18066
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.469005640316674e-05,
+      "loss": 1.937,
+      "step": 18067
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4689526863178854e-05,
+      "loss": 1.9931,
+      "step": 18068
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4688997306333227e-05,
+      "loss": 1.8311,
+      "step": 18069
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4688467732631764e-05,
+      "loss": 2.0573,
+      "step": 18070
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.468793814207637e-05,
+      "loss": 2.0179,
+      "step": 18071
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4687408534668946e-05,
+      "loss": 1.8024,
+      "step": 18072
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4686878910411394e-05,
+      "loss": 2.08,
+      "step": 18073
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4686349269305626e-05,
+      "loss": 1.5899,
+      "step": 18074
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4685819611353537e-05,
+      "loss": 2.1922,
+      "step": 18075
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4685289936557037e-05,
+      "loss": 1.8869,
+      "step": 18076
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4684760244918026e-05,
+      "loss": 1.8663,
+      "step": 18077
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4684230536438411e-05,
+      "loss": 1.8656,
+      "step": 18078
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4683700811120093e-05,
+      "loss": 2.2775,
+      "step": 18079
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.4683171068964978e-05,
+      "loss": 1.7574,
+      "step": 18080
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4682641309974974e-05,
+      "loss": 2.0883,
+      "step": 18081
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.468211153415198e-05,
+      "loss": 2.1286,
+      "step": 18082
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4681581741497902e-05,
+      "loss": 1.7216,
+      "step": 18083
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4681051932014645e-05,
+      "loss": 1.9897,
+      "step": 18084
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4680522105704116e-05,
+      "loss": 1.8653,
+      "step": 18085
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4679992262568213e-05,
+      "loss": 1.7532,
+      "step": 18086
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.4679462402608846e-05,
+      "loss": 2.5613,
+      "step": 18087
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4678932525827918e-05,
+      "loss": 1.7579,
+      "step": 18088
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4678402632227334e-05,
+      "loss": 1.9196,
+      "step": 18089
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4677872721809e-05,
+      "loss": 1.7248,
+      "step": 18090
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4677342794574819e-05,
+      "loss": 2.0614,
+      "step": 18091
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4676812850526698e-05,
+      "loss": 2.1544,
+      "step": 18092
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.467628288966654e-05,
+      "loss": 1.9089,
+      "step": 18093
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.467575291199625e-05,
+      "loss": 1.6461,
+      "step": 18094
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4675222917517735e-05,
+      "loss": 2.1516,
+      "step": 18095
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.46746929062329e-05,
+      "loss": 1.9789,
+      "step": 18096
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.467416287814365e-05,
+      "loss": 1.8015,
+      "step": 18097
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4673632833251886e-05,
+      "loss": 1.7897,
+      "step": 18098
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.467310277155952e-05,
+      "loss": 1.9167,
+      "step": 18099
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4672572693068454e-05,
+      "loss": 2.2203,
+      "step": 18100
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4672042597780596e-05,
+      "loss": 2.0527,
+      "step": 18101
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.467151248569785e-05,
+      "loss": 1.6939,
+      "step": 18102
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.4670982356822117e-05,
+      "loss": 2.2845,
+      "step": 18103
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4670452211155312e-05,
+      "loss": 1.95,
+      "step": 18104
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4669922048699333e-05,
+      "loss": 1.824,
+      "step": 18105
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4669391869456089e-05,
+      "loss": 1.8215,
+      "step": 18106
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4668861673427486e-05,
+      "loss": 1.7341,
+      "step": 18107
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.466833146061543e-05,
+      "loss": 2.1813,
+      "step": 18108
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4667801231021826e-05,
+      "loss": 1.7157,
+      "step": 18109
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4667270984648578e-05,
+      "loss": 2.2023,
+      "step": 18110
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4666740721497598e-05,
+      "loss": 1.8268,
+      "step": 18111
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4666210441570787e-05,
+      "loss": 1.9164,
+      "step": 18112
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4665680144870055e-05,
+      "loss": 2.0188,
+      "step": 18113
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4665149831397307e-05,
+      "loss": 1.7617,
+      "step": 18114
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4664619501154445e-05,
+      "loss": 1.6416,
+      "step": 18115
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4664089154143382e-05,
+      "loss": 1.7342,
+      "step": 18116
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4663558790366018e-05,
+      "loss": 2.0076,
+      "step": 18117
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4663028409824267e-05,
+      "loss": 1.9466,
+      "step": 18118
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4662498012520028e-05,
+      "loss": 1.6927,
+      "step": 18119
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4661967598455214e-05,
+      "loss": 2.1039,
+      "step": 18120
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.466143716763173e-05,
+      "loss": 1.8437,
+      "step": 18121
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4660906720051477e-05,
+      "loss": 1.9735,
+      "step": 18122
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.466037625571637e-05,
+      "loss": 2.0482,
+      "step": 18123
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4659845774628311e-05,
+      "loss": 1.8055,
+      "step": 18124
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.465931527678921e-05,
+      "loss": 2.2869,
+      "step": 18125
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4658784762200972e-05,
+      "loss": 1.7356,
+      "step": 18126
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4658254230865505e-05,
+      "loss": 1.5465,
+      "step": 18127
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4657723682784717e-05,
+      "loss": 1.8451,
+      "step": 18128
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4657193117960509e-05,
+      "loss": 2.0183,
+      "step": 18129
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4656662536394797e-05,
+      "loss": 1.8939,
+      "step": 18130
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4656131938089484e-05,
+      "loss": 1.6199,
+      "step": 18131
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4655601323046477e-05,
+      "loss": 1.9966,
+      "step": 18132
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4655070691267685e-05,
+      "loss": 1.9746,
+      "step": 18133
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4654540042755017e-05,
+      "loss": 1.8648,
+      "step": 18134
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4654009377510374e-05,
+      "loss": 1.8941,
+      "step": 18135
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.465347869553567e-05,
+      "loss": 1.8288,
+      "step": 18136
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.465294799683281e-05,
+      "loss": 1.9121,
+      "step": 18137
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4652417281403702e-05,
+      "loss": 1.5869,
+      "step": 18138
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4651886549250256e-05,
+      "loss": 1.7711,
+      "step": 18139
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4651355800374376e-05,
+      "loss": 2.1384,
+      "step": 18140
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4650825034777974e-05,
+      "loss": 2.0561,
+      "step": 18141
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4650294252462955e-05,
+      "loss": 2.0806,
+      "step": 18142
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4649763453431225e-05,
+      "loss": 1.675,
+      "step": 18143
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4649232637684698e-05,
+      "loss": 1.5279,
+      "step": 18144
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.464870180522528e-05,
+      "loss": 1.5449,
+      "step": 18145
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4648170956054877e-05,
+      "loss": 2.0129,
+      "step": 18146
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.4647640090175398e-05,
+      "loss": 1.499,
+      "step": 18147
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.4647109207588755e-05,
+      "loss": 2.6003,
+      "step": 18148
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.464657830829685e-05,
+      "loss": 1.8388,
+      "step": 18149
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4646047392301594e-05,
+      "loss": 1.9641,
+      "step": 18150
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4645516459604903e-05,
+      "loss": 2.1416,
+      "step": 18151
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4644985510208671e-05,
+      "loss": 1.4904,
+      "step": 18152
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4644454544114818e-05,
+      "loss": 1.7176,
+      "step": 18153
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4643923561325251e-05,
+      "loss": 2.0407,
+      "step": 18154
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4643392561841876e-05,
+      "loss": 1.8205,
+      "step": 18155
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.46428615456666e-05,
+      "loss": 1.7186,
+      "step": 18156
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4642330512801336e-05,
+      "loss": 2.0723,
+      "step": 18157
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4641799463247997e-05,
+      "loss": 2.2307,
+      "step": 18158
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4641268397008478e-05,
+      "loss": 1.5717,
+      "step": 18159
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4640737314084702e-05,
+      "loss": 1.9158,
+      "step": 18160
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4640206214478574e-05,
+      "loss": 1.8281,
+      "step": 18161
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4639675098192e-05,
+      "loss": 1.6037,
+      "step": 18162
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4639143965226889e-05,
+      "loss": 1.9247,
+      "step": 18163
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4638612815585155e-05,
+      "loss": 1.991,
+      "step": 18164
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4638081649268707e-05,
+      "loss": 2.2232,
+      "step": 18165
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4637550466279449e-05,
+      "loss": 1.7837,
+      "step": 18166
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4637019266619295e-05,
+      "loss": 1.6084,
+      "step": 18167
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4636488050290154e-05,
+      "loss": 1.7012,
+      "step": 18168
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4635956817293935e-05,
+      "loss": 2.0007,
+      "step": 18169
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4635425567632547e-05,
+      "loss": 1.638,
+      "step": 18170
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.46348943013079e-05,
+      "loss": 1.7037,
+      "step": 18171
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4634363018321904e-05,
+      "loss": 2.2052,
+      "step": 18172
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4633831718676468e-05,
+      "loss": 1.7665,
+      "step": 18173
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4633300402373504e-05,
+      "loss": 1.8197,
+      "step": 18174
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4632769069414923e-05,
+      "loss": 1.5894,
+      "step": 18175
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.4632237719802629e-05,
+      "loss": 1.8334,
+      "step": 18176
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4631706353538539e-05,
+      "loss": 1.6006,
+      "step": 18177
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4631174970624558e-05,
+      "loss": 1.5653,
+      "step": 18178
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4630643571062599e-05,
+      "loss": 1.7733,
+      "step": 18179
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.463011215485457e-05,
+      "loss": 2.01,
+      "step": 18180
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4629580722002388e-05,
+      "loss": 1.8959,
+      "step": 18181
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4629049272507952e-05,
+      "loss": 1.9942,
+      "step": 18182
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4628517806373181e-05,
+      "loss": 1.6359,
+      "step": 18183
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4627986323599982e-05,
+      "loss": 1.9315,
+      "step": 18184
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4627454824190269e-05,
+      "loss": 2.0294,
+      "step": 18185
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4626923308145948e-05,
+      "loss": 1.6363,
+      "step": 18186
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.462639177546893e-05,
+      "loss": 1.7569,
+      "step": 18187
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4625860226161135e-05,
+      "loss": 2.166,
+      "step": 18188
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4625328660224461e-05,
+      "loss": 2.1034,
+      "step": 18189
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4624797077660823e-05,
+      "loss": 1.8169,
+      "step": 18190
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4624265478472136e-05,
+      "loss": 1.9643,
+      "step": 18191
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.462373386266031e-05,
+      "loss": 1.9411,
+      "step": 18192
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4623202230227251e-05,
+      "loss": 1.984,
+      "step": 18193
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4622670581174873e-05,
+      "loss": 1.9134,
+      "step": 18194
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4622138915505092e-05,
+      "loss": 2.0032,
+      "step": 18195
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.462160723321981e-05,
+      "loss": 2.0042,
+      "step": 18196
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4621075534320946e-05,
+      "loss": 2.1992,
+      "step": 18197
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4620543818810408e-05,
+      "loss": 1.7844,
+      "step": 18198
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4620012086690105e-05,
+      "loss": 1.6872,
+      "step": 18199
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4619480337961953e-05,
+      "loss": 1.9848,
+      "step": 18200
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.461894857262786e-05,
+      "loss": 1.8167,
+      "step": 18201
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4618416790689743e-05,
+      "loss": 1.9005,
+      "step": 18202
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.4617884992149506e-05,
+      "loss": 1.873,
+      "step": 18203
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4617353177009066e-05,
+      "loss": 1.8562,
+      "step": 18204
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4616821345270336e-05,
+      "loss": 1.8709,
+      "step": 18205
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4616289496935219e-05,
+      "loss": 1.7012,
+      "step": 18206
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4615757632005638e-05,
+      "loss": 2.2808,
+      "step": 18207
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4615225750483499e-05,
+      "loss": 1.626,
+      "step": 18208
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4614693852370713e-05,
+      "loss": 1.8154,
+      "step": 18209
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4614161937669196e-05,
+      "loss": 2.0691,
+      "step": 18210
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4613630006380857e-05,
+      "loss": 1.8905,
+      "step": 18211
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.461309805850761e-05,
+      "loss": 1.9031,
+      "step": 18212
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4612566094051364e-05,
+      "loss": 1.5109,
+      "step": 18213
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4612034113014036e-05,
+      "loss": 1.5334,
+      "step": 18214
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4611502115397535e-05,
+      "loss": 2.035,
+      "step": 18215
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4610970101203777e-05,
+      "loss": 2.533,
+      "step": 18216
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.461043807043467e-05,
+      "loss": 1.8573,
+      "step": 18217
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.460990602309213e-05,
+      "loss": 2.0526,
+      "step": 18218
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4609373959178065e-05,
+      "loss": 1.8157,
+      "step": 18219
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4608841878694394e-05,
+      "loss": 1.7809,
+      "step": 18220
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4608309781643022e-05,
+      "loss": 1.8362,
+      "step": 18221
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4607777668025871e-05,
+      "loss": 1.8397,
+      "step": 18222
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4607245537844845e-05,
+      "loss": 2.0454,
+      "step": 18223
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.4606713391101864e-05,
+      "loss": 1.6401,
+      "step": 18224
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4606181227798837e-05,
+      "loss": 1.7579,
+      "step": 18225
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4605649047937677e-05,
+      "loss": 1.9103,
+      "step": 18226
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4605116851520298e-05,
+      "loss": 1.8952,
+      "step": 18227
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4604584638548613e-05,
+      "loss": 1.9022,
+      "step": 18228
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4604052409024536e-05,
+      "loss": 1.8675,
+      "step": 18229
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.460352016294998e-05,
+      "loss": 1.8917,
+      "step": 18230
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.4602987900326854e-05,
+      "loss": 2.0957,
+      "step": 18231
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4602455621157078e-05,
+      "loss": 1.8965,
+      "step": 18232
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4601923325442563e-05,
+      "loss": 2.29,
+      "step": 18233
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.460139101318522e-05,
+      "loss": 2.0641,
+      "step": 18234
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.460085868438697e-05,
+      "loss": 2.0402,
+      "step": 18235
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4600326339049714e-05,
+      "loss": 1.9394,
+      "step": 18236
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4599793977175377e-05,
+      "loss": 1.5719,
+      "step": 18237
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4599261598765865e-05,
+      "loss": 1.8135,
+      "step": 18238
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4598729203823099e-05,
+      "loss": 1.9298,
+      "step": 18239
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4598196792348986e-05,
+      "loss": 1.6056,
+      "step": 18240
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4597664364345443e-05,
+      "loss": 1.9984,
+      "step": 18241
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.459713191981439e-05,
+      "loss": 2.1291,
+      "step": 18242
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4596599458757727e-05,
+      "loss": 1.6379,
+      "step": 18243
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.459606698117738e-05,
+      "loss": 2.0289,
+      "step": 18244
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4595534487075259e-05,
+      "loss": 1.8307,
+      "step": 18245
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4595001976453276e-05,
+      "loss": 2.0934,
+      "step": 18246
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.459446944931335e-05,
+      "loss": 2.0633,
+      "step": 18247
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4593936905657391e-05,
+      "loss": 1.7615,
+      "step": 18248
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4593404345487318e-05,
+      "loss": 1.7797,
+      "step": 18249
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.459287176880504e-05,
+      "loss": 1.6267,
+      "step": 18250
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4592339175612475e-05,
+      "loss": 1.7781,
+      "step": 18251
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4591806565911538e-05,
+      "loss": 2.0468,
+      "step": 18252
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4591273939704142e-05,
+      "loss": 1.9944,
+      "step": 18253
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.45907412969922e-05,
+      "loss": 1.8564,
+      "step": 18254
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4590208637777632e-05,
+      "loss": 1.8745,
+      "step": 18255
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4589675962062346e-05,
+      "loss": 2.48,
+      "step": 18256
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4589143269848261e-05,
+      "loss": 1.9184,
+      "step": 18257
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.4588610561137293e-05,
+      "loss": 2.0323,
+      "step": 18258
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4588077835931353e-05,
+      "loss": 1.9355,
+      "step": 18259
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.458754509423236e-05,
+      "loss": 1.5146,
+      "step": 18260
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4587012336042228e-05,
+      "loss": 1.9877,
+      "step": 18261
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4586479561362872e-05,
+      "loss": 1.7354,
+      "step": 18262
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4585946770196205e-05,
+      "loss": 2.0668,
+      "step": 18263
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4585413962544141e-05,
+      "loss": 1.9626,
+      "step": 18264
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4584881138408603e-05,
+      "loss": 2.0356,
+      "step": 18265
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.45843482977915e-05,
+      "loss": 1.9048,
+      "step": 18266
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4583815440694746e-05,
+      "loss": 1.7061,
+      "step": 18267
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4583282567120263e-05,
+      "loss": 2.0166,
+      "step": 18268
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4582749677069963e-05,
+      "loss": 1.8311,
+      "step": 18269
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4582216770545762e-05,
+      "loss": 1.9563,
+      "step": 18270
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4581683847549573e-05,
+      "loss": 2.004,
+      "step": 18271
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4581150908083316e-05,
+      "loss": 1.5964,
+      "step": 18272
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4580617952148903e-05,
+      "loss": 2.0908,
+      "step": 18273
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4580084979748254e-05,
+      "loss": 2.0504,
+      "step": 18274
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4579551990883283e-05,
+      "loss": 1.4166,
+      "step": 18275
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4579018985555904e-05,
+      "loss": 1.9356,
+      "step": 18276
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4578485963768036e-05,
+      "loss": 1.9518,
+      "step": 18277
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.457795292552159e-05,
+      "loss": 1.9676,
+      "step": 18278
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4577419870818492e-05,
+      "loss": 1.9975,
+      "step": 18279
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4576886799660647e-05,
+      "loss": 1.9769,
+      "step": 18280
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4576353712049978e-05,
+      "loss": 1.8295,
+      "step": 18281
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4575820607988403e-05,
+      "loss": 2.0302,
+      "step": 18282
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4575287487477829e-05,
+      "loss": 2.0845,
+      "step": 18283
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.4574754350520181e-05,
+      "loss": 1.391,
+      "step": 18284
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4574221197117373e-05,
+      "loss": 1.8968,
+      "step": 18285
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4573688027271325e-05,
+      "loss": 2.2711,
+      "step": 18286
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4573154840983947e-05,
+      "loss": 1.7481,
+      "step": 18287
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4572621638257158e-05,
+      "loss": 2.0542,
+      "step": 18288
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4572088419092877e-05,
+      "loss": 1.7154,
+      "step": 18289
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4571555183493018e-05,
+      "loss": 1.7112,
+      "step": 18290
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4571021931459502e-05,
+      "loss": 1.6986,
+      "step": 18291
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4570488662994242e-05,
+      "loss": 2.0147,
+      "step": 18292
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4569955378099154e-05,
+      "loss": 2.212,
+      "step": 18293
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.456942207677616e-05,
+      "loss": 1.5757,
+      "step": 18294
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.456888875902717e-05,
+      "loss": 1.7747,
+      "step": 18295
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4568355424854113e-05,
+      "loss": 1.6693,
+      "step": 18296
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.456782207425889e-05,
+      "loss": 1.7574,
+      "step": 18297
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4567288707243435e-05,
+      "loss": 1.9279,
+      "step": 18298
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4566755323809652e-05,
+      "loss": 1.6279,
+      "step": 18299
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4566221923959465e-05,
+      "loss": 1.8276,
+      "step": 18300
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.456568850769479e-05,
+      "loss": 2.0714,
+      "step": 18301
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4565155075017545e-05,
+      "loss": 1.7864,
+      "step": 18302
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4564621625929647e-05,
+      "loss": 1.8627,
+      "step": 18303
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4564088160433011e-05,
+      "loss": 1.9754,
+      "step": 18304
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.456355467852956e-05,
+      "loss": 1.8528,
+      "step": 18305
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4563021180221211e-05,
+      "loss": 1.6995,
+      "step": 18306
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4562487665509876e-05,
+      "loss": 1.6744,
+      "step": 18307
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.456195413439748e-05,
+      "loss": 2.0801,
+      "step": 18308
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4561420586885936e-05,
+      "loss": 1.7177,
+      "step": 18309
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4560887022977166e-05,
+      "loss": 2.0991,
+      "step": 18310
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4560353442673083e-05,
+      "loss": 1.7693,
+      "step": 18311
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4559819845975609e-05,
+      "loss": 1.8728,
+      "step": 18312
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4559286232886662e-05,
+      "loss": 1.9138,
+      "step": 18313
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4558752603408158e-05,
+      "loss": 2.1847,
+      "step": 18314
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4558218957542017e-05,
+      "loss": 1.9831,
+      "step": 18315
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4557685295290156e-05,
+      "loss": 1.6653,
+      "step": 18316
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4557151616654496e-05,
+      "loss": 1.7704,
+      "step": 18317
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.455661792163695e-05,
+      "loss": 1.7699,
+      "step": 18318
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4556084210239445e-05,
+      "loss": 1.5405,
+      "step": 18319
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4555550482463892e-05,
+      "loss": 1.9482,
+      "step": 18320
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.455501673831221e-05,
+      "loss": 1.6413,
+      "step": 18321
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4554482977786327e-05,
+      "loss": 1.7496,
+      "step": 18322
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4553949200888148e-05,
+      "loss": 1.7035,
+      "step": 18323
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4553415407619603e-05,
+      "loss": 1.7439,
+      "step": 18324
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4552881597982602e-05,
+      "loss": 2.1638,
+      "step": 18325
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4552347771979074e-05,
+      "loss": 1.8205,
+      "step": 18326
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4551813929610927e-05,
+      "loss": 1.9775,
+      "step": 18327
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4551280070880089e-05,
+      "loss": 1.79,
+      "step": 18328
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4550746195788474e-05,
+      "loss": 1.921,
+      "step": 18329
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4550212304338006e-05,
+      "loss": 1.8665,
+      "step": 18330
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.4549678396530596e-05,
+      "loss": 2.2972,
+      "step": 18331
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4549144472368169e-05,
+      "loss": 1.84,
+      "step": 18332
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4548610531852646e-05,
+      "loss": 1.9388,
+      "step": 18333
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.4548076574985941e-05,
+      "loss": 2.1236,
+      "step": 18334
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4547542601769977e-05,
+      "loss": 1.9035,
+      "step": 18335
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4547008612206674e-05,
+      "loss": 1.9147,
+      "step": 18336
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4546474606297951e-05,
+      "loss": 1.8291,
+      "step": 18337
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4545940584045726e-05,
+      "loss": 1.8959,
+      "step": 18338
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.454540654545192e-05,
+      "loss": 1.6439,
+      "step": 18339
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.454487249051845e-05,
+      "loss": 2.0109,
+      "step": 18340
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.4544338419247239e-05,
+      "loss": 2.083,
+      "step": 18341
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.4543804331640208e-05,
+      "loss": 1.2573,
+      "step": 18342
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4543270227699275e-05,
+      "loss": 2.0714,
+      "step": 18343
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4542736107426359e-05,
+      "loss": 1.8804,
+      "step": 18344
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4542201970823383e-05,
+      "loss": 2.1083,
+      "step": 18345
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4541667817892262e-05,
+      "loss": 1.6196,
+      "step": 18346
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4541133648634921e-05,
+      "loss": 1.8115,
+      "step": 18347
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4540599463053278e-05,
+      "loss": 1.6644,
+      "step": 18348
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4540065261149254e-05,
+      "loss": 1.791,
+      "step": 18349
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4539531042924769e-05,
+      "loss": 2.1514,
+      "step": 18350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4538996808381743e-05,
+      "loss": 2.0781,
+      "step": 18351
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4538462557522098e-05,
+      "loss": 2.0504,
+      "step": 18352
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4537928290347751e-05,
+      "loss": 1.909,
+      "step": 18353
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4537394006860627e-05,
+      "loss": 1.8514,
+      "step": 18354
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.4536859707062644e-05,
+      "loss": 2.2936,
+      "step": 18355
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4536325390955726e-05,
+      "loss": 2.1166,
+      "step": 18356
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4535791058541789e-05,
+      "loss": 1.7913,
+      "step": 18357
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4535256709822754e-05,
+      "loss": 1.8143,
+      "step": 18358
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4534722344800545e-05,
+      "loss": 1.8328,
+      "step": 18359
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4534187963477082e-05,
+      "loss": 2.0439,
+      "step": 18360
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4533653565854286e-05,
+      "loss": 2.1183,
+      "step": 18361
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4533119151934077e-05,
+      "loss": 1.5909,
+      "step": 18362
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4532584721718377e-05,
+      "loss": 1.5999,
+      "step": 18363
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4532050275209106e-05,
+      "loss": 1.8313,
+      "step": 18364
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4531515812408186e-05,
+      "loss": 1.8494,
+      "step": 18365
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.453098133331754e-05,
+      "loss": 1.9613,
+      "step": 18366
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4530446837939084e-05,
+      "loss": 1.6452,
+      "step": 18367
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4529912326274746e-05,
+      "loss": 1.5908,
+      "step": 18368
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.4529377798326443e-05,
+      "loss": 1.8224,
+      "step": 18369
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4528843254096101e-05,
+      "loss": 1.626,
+      "step": 18370
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4528308693585637e-05,
+      "loss": 1.7138,
+      "step": 18371
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4527774116796974e-05,
+      "loss": 1.8684,
+      "step": 18372
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4527239523732035e-05,
+      "loss": 1.8331,
+      "step": 18373
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4526704914392738e-05,
+      "loss": 2.2726,
+      "step": 18374
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4526170288781009e-05,
+      "loss": 1.8905,
+      "step": 18375
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4525635646898768e-05,
+      "loss": 1.6078,
+      "step": 18376
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4525100988747935e-05,
+      "loss": 1.8461,
+      "step": 18377
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4524566314330439e-05,
+      "loss": 1.9868,
+      "step": 18378
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4524031623648192e-05,
+      "loss": 1.4779,
+      "step": 18379
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4523496916703126e-05,
+      "loss": 1.7046,
+      "step": 18380
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4522962193497154e-05,
+      "loss": 1.8751,
+      "step": 18381
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4522427454032203e-05,
+      "loss": 1.9008,
+      "step": 18382
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4521892698310196e-05,
+      "loss": 1.8099,
+      "step": 18383
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4521357926333055e-05,
+      "loss": 1.9635,
+      "step": 18384
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4520823138102703e-05,
+      "loss": 1.7758,
+      "step": 18385
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4520288333621058e-05,
+      "loss": 1.6757,
+      "step": 18386
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4519753512890046e-05,
+      "loss": 2.0084,
+      "step": 18387
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4519218675911588e-05,
+      "loss": 1.6402,
+      "step": 18388
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.451868382268761e-05,
+      "loss": 1.4703,
+      "step": 18389
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.451814895322003e-05,
+      "loss": 1.8613,
+      "step": 18390
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4517614067510775e-05,
+      "loss": 1.8185,
+      "step": 18391
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4517079165561764e-05,
+      "loss": 1.7297,
+      "step": 18392
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4516544247374925e-05,
+      "loss": 2.1179,
+      "step": 18393
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4516009312952175e-05,
+      "loss": 1.9368,
+      "step": 18394
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4515474362295437e-05,
+      "loss": 1.9762,
+      "step": 18395
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4514939395406642e-05,
+      "loss": 1.8133,
+      "step": 18396
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4514404412287704e-05,
+      "loss": 1.6607,
+      "step": 18397
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.451386941294055e-05,
+      "loss": 1.9367,
+      "step": 18398
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4513334397367103e-05,
+      "loss": 1.6847,
+      "step": 18399
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4512799365569288e-05,
+      "loss": 1.8206,
+      "step": 18400
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4512264317549025e-05,
+      "loss": 1.8368,
+      "step": 18401
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4511729253308238e-05,
+      "loss": 1.6455,
+      "step": 18402
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4511194172848855e-05,
+      "loss": 1.7848,
+      "step": 18403
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4510659076172791e-05,
+      "loss": 1.8564,
+      "step": 18404
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4510123963281976e-05,
+      "loss": 2.0742,
+      "step": 18405
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4509588834178334e-05,
+      "loss": 1.9529,
+      "step": 18406
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4509053688863786e-05,
+      "loss": 2.0973,
+      "step": 18407
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4508518527340256e-05,
+      "loss": 2.1751,
+      "step": 18408
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4507983349609667e-05,
+      "loss": 2.0089,
+      "step": 18409
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4507448155673947e-05,
+      "loss": 2.1662,
+      "step": 18410
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4506912945535013e-05,
+      "loss": 2.0507,
+      "step": 18411
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4506377719194794e-05,
+      "loss": 1.9631,
+      "step": 18412
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4505842476655212e-05,
+      "loss": 1.8084,
+      "step": 18413
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4505307217918195e-05,
+      "loss": 1.8269,
+      "step": 18414
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4504771942985662e-05,
+      "loss": 1.7274,
+      "step": 18415
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4504236651859539e-05,
+      "loss": 1.6635,
+      "step": 18416
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4503701344541752e-05,
+      "loss": 1.9235,
+      "step": 18417
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.450316602103422e-05,
+      "loss": 1.8098,
+      "step": 18418
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.4502630681338876e-05,
+      "loss": 1.9817,
+      "step": 18419
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4502095325457637e-05,
+      "loss": 1.9233,
+      "step": 18420
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.450155995339243e-05,
+      "loss": 2.0587,
+      "step": 18421
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4501024565145181e-05,
+      "loss": 1.8671,
+      "step": 18422
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4500489160717813e-05,
+      "loss": 1.6541,
+      "step": 18423
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.449995374011225e-05,
+      "loss": 1.7884,
+      "step": 18424
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4499418303330414e-05,
+      "loss": 1.8234,
+      "step": 18425
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4498882850374237e-05,
+      "loss": 1.938,
+      "step": 18426
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4498347381245642e-05,
+      "loss": 1.9152,
+      "step": 18427
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.4497811895946549e-05,
+      "loss": 1.6409,
+      "step": 18428
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4497276394478887e-05,
+      "loss": 1.9753,
+      "step": 18429
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.449674087684458e-05,
+      "loss": 1.8135,
+      "step": 18430
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4496205343045552e-05,
+      "loss": 1.6958,
+      "step": 18431
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.449566979308373e-05,
+      "loss": 1.9499,
+      "step": 18432
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4495134226961039e-05,
+      "loss": 1.95,
+      "step": 18433
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.44945986446794e-05,
+      "loss": 1.7971,
+      "step": 18434
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4494063046240744e-05,
+      "loss": 1.6764,
+      "step": 18435
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.4493527431646997e-05,
+      "loss": 2.0964,
+      "step": 18436
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4492991800900077e-05,
+      "loss": 2.0671,
+      "step": 18437
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4492456154001917e-05,
+      "loss": 1.7813,
+      "step": 18438
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4491920490954439e-05,
+      "loss": 1.6517,
+      "step": 18439
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.449138481175957e-05,
+      "loss": 1.8913,
+      "step": 18440
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4490849116419231e-05,
+      "loss": 2.2324,
+      "step": 18441
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4490313404935354e-05,
+      "loss": 1.8129,
+      "step": 18442
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.625,
+      "learning_rate": 1.4489777677309864e-05,
+      "loss": 3.2435,
+      "step": 18443
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4489241933544685e-05,
+      "loss": 1.8708,
+      "step": 18444
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.448870617364174e-05,
+      "loss": 1.7986,
+      "step": 18445
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4488170397602957e-05,
+      "loss": 1.7503,
+      "step": 18446
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4487634605430268e-05,
+      "loss": 1.9289,
+      "step": 18447
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4487098797125592e-05,
+      "loss": 1.7552,
+      "step": 18448
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.4486562972690851e-05,
+      "loss": 2.3353,
+      "step": 18449
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4486027132127987e-05,
+      "loss": 1.9019,
+      "step": 18450
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.448549127543891e-05,
+      "loss": 1.8665,
+      "step": 18451
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4484955402625554e-05,
+      "loss": 2.2361,
+      "step": 18452
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4484419513689843e-05,
+      "loss": 1.8422,
+      "step": 18453
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.4483883608633707e-05,
+      "loss": 1.9652,
+      "step": 18454
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4483347687459068e-05,
+      "loss": 1.9109,
+      "step": 18455
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4482811750167853e-05,
+      "loss": 1.7639,
+      "step": 18456
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4482275796761993e-05,
+      "loss": 1.774,
+      "step": 18457
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4481739827243409e-05,
+      "loss": 2.264,
+      "step": 18458
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4481203841614031e-05,
+      "loss": 1.7808,
+      "step": 18459
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4480667839875786e-05,
+      "loss": 1.92,
+      "step": 18460
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4480131822030598e-05,
+      "loss": 1.6976,
+      "step": 18461
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4479595788080398e-05,
+      "loss": 2.1112,
+      "step": 18462
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4479059738027108e-05,
+      "loss": 2.0071,
+      "step": 18463
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4478523671872661e-05,
+      "loss": 2.0907,
+      "step": 18464
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4477987589618976e-05,
+      "loss": 1.726,
+      "step": 18465
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4477451491267989e-05,
+      "loss": 1.7726,
+      "step": 18466
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4476915376821619e-05,
+      "loss": 1.8819,
+      "step": 18467
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.44763792462818e-05,
+      "loss": 1.7021,
+      "step": 18468
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4475843099650455e-05,
+      "loss": 2.2166,
+      "step": 18469
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4475306936929513e-05,
+      "loss": 1.8383,
+      "step": 18470
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.44747707581209e-05,
+      "loss": 1.8018,
+      "step": 18471
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4474234563226543e-05,
+      "loss": 1.7538,
+      "step": 18472
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4473698352248373e-05,
+      "loss": 1.9529,
+      "step": 18473
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4473162125188315e-05,
+      "loss": 2.1844,
+      "step": 18474
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4472625882048298e-05,
+      "loss": 1.9339,
+      "step": 18475
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4472089622830248e-05,
+      "loss": 1.7196,
+      "step": 18476
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4471553347536094e-05,
+      "loss": 1.9102,
+      "step": 18477
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4471017056167762e-05,
+      "loss": 1.925,
+      "step": 18478
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4470480748727181e-05,
+      "loss": 1.7552,
+      "step": 18479
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.446994442521628e-05,
+      "loss": 2.0245,
+      "step": 18480
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4469408085636985e-05,
+      "loss": 1.9933,
+      "step": 18481
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4468871729991226e-05,
+      "loss": 1.897,
+      "step": 18482
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.446833535828093e-05,
+      "loss": 1.8491,
+      "step": 18483
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4467798970508022e-05,
+      "loss": 2.0448,
+      "step": 18484
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4467262566674436e-05,
+      "loss": 1.9747,
+      "step": 18485
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4466726146782098e-05,
+      "loss": 1.6642,
+      "step": 18486
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4466189710832935e-05,
+      "loss": 2.0455,
+      "step": 18487
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4465653258828875e-05,
+      "loss": 1.8714,
+      "step": 18488
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4465116790771849e-05,
+      "loss": 1.8634,
+      "step": 18489
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4464580306663785e-05,
+      "loss": 1.7808,
+      "step": 18490
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4464043806506609e-05,
+      "loss": 1.8955,
+      "step": 18491
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4463507290302252e-05,
+      "loss": 1.9602,
+      "step": 18492
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.446297075805264e-05,
+      "loss": 2.0991,
+      "step": 18493
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.4462434209759709e-05,
+      "loss": 1.9182,
+      "step": 18494
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4461897645425375e-05,
+      "loss": 1.834,
+      "step": 18495
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.446136106505158e-05,
+      "loss": 2.2139,
+      "step": 18496
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4460824468640244e-05,
+      "loss": 2.1031,
+      "step": 18497
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.44602878561933e-05,
+      "loss": 1.9158,
+      "step": 18498
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4459751227712678e-05,
+      "loss": 1.8708,
+      "step": 18499
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4459214583200301e-05,
+      "loss": 1.7467,
+      "step": 18500
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4458677922658107e-05,
+      "loss": 1.682,
+      "step": 18501
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4458141246088015e-05,
+      "loss": 1.6444,
+      "step": 18502
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4457604553491961e-05,
+      "loss": 1.7654,
+      "step": 18503
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4457067844871875e-05,
+      "loss": 2.0766,
+      "step": 18504
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4456531120229682e-05,
+      "loss": 2.294,
+      "step": 18505
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4455994379567316e-05,
+      "loss": 1.9392,
+      "step": 18506
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4455457622886701e-05,
+      "loss": 1.9691,
+      "step": 18507
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.4454920850189774e-05,
+      "loss": 1.9868,
+      "step": 18508
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4454384061478454e-05,
+      "loss": 1.8292,
+      "step": 18509
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.445384725675468e-05,
+      "loss": 1.6742,
+      "step": 18510
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4453310436020377e-05,
+      "loss": 1.9731,
+      "step": 18511
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4452773599277478e-05,
+      "loss": 1.5891,
+      "step": 18512
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.4452236746527913e-05,
+      "loss": 2.003,
+      "step": 18513
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4451699877773606e-05,
+      "loss": 1.4456,
+      "step": 18514
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4451162993016491e-05,
+      "loss": 1.8626,
+      "step": 18515
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.44506260922585e-05,
+      "loss": 2.042,
+      "step": 18516
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.445008917550156e-05,
+      "loss": 1.7212,
+      "step": 18517
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.44495522427476e-05,
+      "loss": 1.6493,
+      "step": 18518
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4449015293998554e-05,
+      "loss": 1.9299,
+      "step": 18519
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.444847832925635e-05,
+      "loss": 1.7465,
+      "step": 18520
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4447941348522917e-05,
+      "loss": 1.9032,
+      "step": 18521
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4447404351800187e-05,
+      "loss": 1.9287,
+      "step": 18522
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4446867339090091e-05,
+      "loss": 1.8407,
+      "step": 18523
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.444633031039456e-05,
+      "loss": 1.5834,
+      "step": 18524
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.444579326571552e-05,
+      "loss": 1.9352,
+      "step": 18525
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4445256205054905e-05,
+      "loss": 1.6041,
+      "step": 18526
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4444719128414648e-05,
+      "loss": 1.7441,
+      "step": 18527
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4444182035796675e-05,
+      "loss": 1.8139,
+      "step": 18528
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4443644927202917e-05,
+      "loss": 2.0932,
+      "step": 18529
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4443107802635309e-05,
+      "loss": 1.7568,
+      "step": 18530
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.444257066209578e-05,
+      "loss": 1.5869,
+      "step": 18531
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4442033505586257e-05,
+      "loss": 1.9697,
+      "step": 18532
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4441496333108676e-05,
+      "loss": 1.8736,
+      "step": 18533
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4440959144664968e-05,
+      "loss": 1.9073,
+      "step": 18534
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4440421940257057e-05,
+      "loss": 1.9363,
+      "step": 18535
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4439884719886883e-05,
+      "loss": 1.8272,
+      "step": 18536
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4439347483556369e-05,
+      "loss": 1.8684,
+      "step": 18537
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4438810231267458e-05,
+      "loss": 1.8381,
+      "step": 18538
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4438272963022068e-05,
+      "loss": 1.7818,
+      "step": 18539
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4437735678822138e-05,
+      "loss": 2.0278,
+      "step": 18540
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4437198378669598e-05,
+      "loss": 1.8217,
+      "step": 18541
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.443666106256638e-05,
+      "loss": 2.0129,
+      "step": 18542
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4436123730514412e-05,
+      "loss": 1.6224,
+      "step": 18543
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4435586382515632e-05,
+      "loss": 1.8336,
+      "step": 18544
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4435049018571967e-05,
+      "loss": 1.9494,
+      "step": 18545
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4434511638685348e-05,
+      "loss": 2.5777,
+      "step": 18546
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4433974242857707e-05,
+      "loss": 2.0543,
+      "step": 18547
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4433436831090984e-05,
+      "loss": 1.9344,
+      "step": 18548
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4432899403387097e-05,
+      "loss": 2.0063,
+      "step": 18549
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4432361959747987e-05,
+      "loss": 1.9332,
+      "step": 18550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4431824500175587e-05,
+      "loss": 2.0203,
+      "step": 18551
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4431287024671821e-05,
+      "loss": 1.8287,
+      "step": 18552
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4430749533238631e-05,
+      "loss": 1.9274,
+      "step": 18553
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4430212025877941e-05,
+      "loss": 1.7498,
+      "step": 18554
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4429674502591686e-05,
+      "loss": 1.8117,
+      "step": 18555
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.44291369633818e-05,
+      "loss": 1.9352,
+      "step": 18556
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4428599408250214e-05,
+      "loss": 1.9394,
+      "step": 18557
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.442806183719886e-05,
+      "loss": 1.5931,
+      "step": 18558
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4427524250229671e-05,
+      "loss": 1.9337,
+      "step": 18559
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4426986647344584e-05,
+      "loss": 2.1519,
+      "step": 18560
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4426449028545522e-05,
+      "loss": 1.8731,
+      "step": 18561
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4425911393834424e-05,
+      "loss": 1.5877,
+      "step": 18562
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4425373743213219e-05,
+      "loss": 1.7615,
+      "step": 18563
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4424836076683846e-05,
+      "loss": 2.2305,
+      "step": 18564
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.4424298394248229e-05,
+      "loss": 2.1404,
+      "step": 18565
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4423760695908308e-05,
+      "loss": 1.5334,
+      "step": 18566
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4423222981666015e-05,
+      "loss": 1.9379,
+      "step": 18567
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4422685251523278e-05,
+      "loss": 2.0674,
+      "step": 18568
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.4422147505482037e-05,
+      "loss": 1.8558,
+      "step": 18569
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4421609743544217e-05,
+      "loss": 1.6686,
+      "step": 18570
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.442107196571176e-05,
+      "loss": 1.9619,
+      "step": 18571
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4420534171986592e-05,
+      "loss": 2.0583,
+      "step": 18572
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4419996362370652e-05,
+      "loss": 1.7132,
+      "step": 18573
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4419458536865869e-05,
+      "loss": 1.8549,
+      "step": 18574
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4418920695474179e-05,
+      "loss": 1.9858,
+      "step": 18575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.441838283819751e-05,
+      "loss": 1.8519,
+      "step": 18576
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4417844965037801e-05,
+      "loss": 1.4555,
+      "step": 18577
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.441730707599699e-05,
+      "loss": 1.8193,
+      "step": 18578
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4416769171076999e-05,
+      "loss": 1.8826,
+      "step": 18579
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4416231250279768e-05,
+      "loss": 1.8923,
+      "step": 18580
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.441569331360723e-05,
+      "loss": 2.0283,
+      "step": 18581
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.441515536106132e-05,
+      "loss": 2.1259,
+      "step": 18582
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.441461739264397e-05,
+      "loss": 1.7759,
+      "step": 18583
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4414079408357114e-05,
+      "loss": 2.149,
+      "step": 18584
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.441354140820269e-05,
+      "loss": 2.1926,
+      "step": 18585
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4413003392182623e-05,
+      "loss": 2.2739,
+      "step": 18586
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4412465360298855e-05,
+      "loss": 2.288,
+      "step": 18587
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4411927312553318e-05,
+      "loss": 2.1522,
+      "step": 18588
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4411389248947944e-05,
+      "loss": 1.5522,
+      "step": 18589
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4410851169484673e-05,
+      "loss": 1.6216,
+      "step": 18590
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4410313074165431e-05,
+      "loss": 2.1716,
+      "step": 18591
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4409774962992159e-05,
+      "loss": 1.9334,
+      "step": 18592
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4409236835966785e-05,
+      "loss": 1.8058,
+      "step": 18593
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.440869869309125e-05,
+      "loss": 1.7859,
+      "step": 18594
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4408160534367485e-05,
+      "loss": 1.6005,
+      "step": 18595
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4407622359797428e-05,
+      "loss": 1.9422,
+      "step": 18596
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.440708416938301e-05,
+      "loss": 1.742,
+      "step": 18597
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4406545963126166e-05,
+      "loss": 1.9558,
+      "step": 18598
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4406007741028832e-05,
+      "loss": 1.9695,
+      "step": 18599
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.4405469503092939e-05,
+      "loss": 1.9874,
+      "step": 18600
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.440493124932043e-05,
+      "loss": 1.723,
+      "step": 18601
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.4404392979713229e-05,
+      "loss": 2.3729,
+      "step": 18602
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4403854694273281e-05,
+      "loss": 1.8221,
+      "step": 18603
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4403316393002515e-05,
+      "loss": 1.9369,
+      "step": 18604
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.440277807590287e-05,
+      "loss": 2.035,
+      "step": 18605
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4402239742976279e-05,
+      "loss": 1.9993,
+      "step": 18606
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4401701394224674e-05,
+      "loss": 2.0193,
+      "step": 18607
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4401163029649998e-05,
+      "loss": 2.182,
+      "step": 18608
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4400624649254177e-05,
+      "loss": 2.0083,
+      "step": 18609
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4400086253039155e-05,
+      "loss": 1.5503,
+      "step": 18610
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4399547841006862e-05,
+      "loss": 1.928,
+      "step": 18611
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4399009413159234e-05,
+      "loss": 1.8757,
+      "step": 18612
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.439847096949821e-05,
+      "loss": 1.6556,
+      "step": 18613
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4397932510025721e-05,
+      "loss": 1.8404,
+      "step": 18614
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4397394034743707e-05,
+      "loss": 1.9533,
+      "step": 18615
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4396855543654099e-05,
+      "loss": 1.7799,
+      "step": 18616
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4396317036758834e-05,
+      "loss": 1.8604,
+      "step": 18617
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4395778514059852e-05,
+      "loss": 1.8767,
+      "step": 18618
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4395239975559083e-05,
+      "loss": 2.2068,
+      "step": 18619
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4394701421258467e-05,
+      "loss": 1.8942,
+      "step": 18620
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4394162851159939e-05,
+      "loss": 2.053,
+      "step": 18621
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4393624265265436e-05,
+      "loss": 1.629,
+      "step": 18622
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4393085663576893e-05,
+      "loss": 1.7937,
+      "step": 18623
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4392547046096241e-05,
+      "loss": 1.7762,
+      "step": 18624
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.439200841282543e-05,
+      "loss": 2.04,
+      "step": 18625
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.439146976376638e-05,
+      "loss": 1.9563,
+      "step": 18626
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4390931098921035e-05,
+      "loss": 1.8434,
+      "step": 18627
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4390392418291334e-05,
+      "loss": 1.9256,
+      "step": 18628
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.438985372187921e-05,
+      "loss": 2.0132,
+      "step": 18629
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.43893150096866e-05,
+      "loss": 1.5579,
+      "step": 18630
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4388776281715438e-05,
+      "loss": 1.8606,
+      "step": 18631
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4388237537967669e-05,
+      "loss": 1.9427,
+      "step": 18632
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4387698778445219e-05,
+      "loss": 1.9361,
+      "step": 18633
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.438716000315003e-05,
+      "loss": 1.7684,
+      "step": 18634
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4386621212084037e-05,
+      "loss": 1.4805,
+      "step": 18635
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4386082405249182e-05,
+      "loss": 1.7117,
+      "step": 18636
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4385543582647397e-05,
+      "loss": 1.9566,
+      "step": 18637
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.438500474428062e-05,
+      "loss": 2.0161,
+      "step": 18638
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4384465890150787e-05,
+      "loss": 2.2244,
+      "step": 18639
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4383927020259836e-05,
+      "loss": 1.8141,
+      "step": 18640
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4383388134609704e-05,
+      "loss": 2.108,
+      "step": 18641
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4382849233202327e-05,
+      "loss": 1.8572,
+      "step": 18642
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4382310316039646e-05,
+      "loss": 1.8439,
+      "step": 18643
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4381771383123596e-05,
+      "loss": 1.7297,
+      "step": 18644
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4381232434456114e-05,
+      "loss": 2.0357,
+      "step": 18645
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4380693470039135e-05,
+      "loss": 2.0174,
+      "step": 18646
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4380154489874601e-05,
+      "loss": 2.1855,
+      "step": 18647
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.437961549396445e-05,
+      "loss": 1.7324,
+      "step": 18648
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.437907648231061e-05,
+      "loss": 1.7663,
+      "step": 18649
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4378537454915032e-05,
+      "loss": 2.0489,
+      "step": 18650
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4377998411779643e-05,
+      "loss": 1.9132,
+      "step": 18651
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.437745935290639e-05,
+      "loss": 1.6999,
+      "step": 18652
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4376920278297202e-05,
+      "loss": 2.0022,
+      "step": 18653
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.437638118795402e-05,
+      "loss": 1.9819,
+      "step": 18654
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4375842081878788e-05,
+      "loss": 1.8569,
+      "step": 18655
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4375302960073434e-05,
+      "loss": 1.7302,
+      "step": 18656
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.43747638225399e-05,
+      "loss": 2.167,
+      "step": 18657
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4374224669280126e-05,
+      "loss": 1.5908,
+      "step": 18658
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4373685500296051e-05,
+      "loss": 1.8912,
+      "step": 18659
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.437314631558961e-05,
+      "loss": 1.8619,
+      "step": 18660
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4372607115162738e-05,
+      "loss": 1.5004,
+      "step": 18661
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4372067899017384e-05,
+      "loss": 1.6099,
+      "step": 18662
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4371528667155475e-05,
+      "loss": 2.0224,
+      "step": 18663
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4370989419578956e-05,
+      "loss": 1.771,
+      "step": 18664
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.4370450156289763e-05,
+      "loss": 1.7703,
+      "step": 18665
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4369910877289837e-05,
+      "loss": 1.738,
+      "step": 18666
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4369371582581113e-05,
+      "loss": 1.8998,
+      "step": 18667
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4368832272165531e-05,
+      "loss": 1.7314,
+      "step": 18668
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4368292946045035e-05,
+      "loss": 2.0932,
+      "step": 18669
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4367753604221552e-05,
+      "loss": 1.7824,
+      "step": 18670
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4367214246697032e-05,
+      "loss": 1.6972,
+      "step": 18671
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4366674873473409e-05,
+      "loss": 2.0464,
+      "step": 18672
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.4366135484552621e-05,
+      "loss": 2.1451,
+      "step": 18673
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4365596079936613e-05,
+      "loss": 1.5664,
+      "step": 18674
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4365056659627315e-05,
+      "loss": 1.9103,
+      "step": 18675
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4364517223626672e-05,
+      "loss": 1.922,
+      "step": 18676
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.436397777193662e-05,
+      "loss": 1.617,
+      "step": 18677
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4363438304559103e-05,
+      "loss": 1.7611,
+      "step": 18678
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4362898821496055e-05,
+      "loss": 1.8758,
+      "step": 18679
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.4362359322749417e-05,
+      "loss": 1.962,
+      "step": 18680
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4361819808321133e-05,
+      "loss": 1.5906,
+      "step": 18681
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4361280278213135e-05,
+      "loss": 1.9033,
+      "step": 18682
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4360740732427367e-05,
+      "loss": 1.8919,
+      "step": 18683
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4360201170965763e-05,
+      "loss": 1.8181,
+      "step": 18684
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4359661593830273e-05,
+      "loss": 1.8574,
+      "step": 18685
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.435912200102283e-05,
+      "loss": 1.888,
+      "step": 18686
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.435858239254537e-05,
+      "loss": 1.9427,
+      "step": 18687
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.435804276839984e-05,
+      "loss": 2.1224,
+      "step": 18688
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4357503128588178e-05,
+      "loss": 1.7384,
+      "step": 18689
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4356963473112322e-05,
+      "loss": 1.9834,
+      "step": 18690
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4356423801974211e-05,
+      "loss": 1.9604,
+      "step": 18691
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4355884115175788e-05,
+      "loss": 1.8674,
+      "step": 18692
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4355344412718992e-05,
+      "loss": 1.9265,
+      "step": 18693
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4354804694605761e-05,
+      "loss": 1.7595,
+      "step": 18694
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.435426496083804e-05,
+      "loss": 1.5621,
+      "step": 18695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4353725211417765e-05,
+      "loss": 2.0878,
+      "step": 18696
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4353185446346875e-05,
+      "loss": 1.9729,
+      "step": 18697
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4352645665627315e-05,
+      "loss": 1.9779,
+      "step": 18698
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4352105869261026e-05,
+      "loss": 1.8021,
+      "step": 18699
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.435156605724994e-05,
+      "loss": 2.086,
+      "step": 18700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4351026229596005e-05,
+      "loss": 1.8871,
+      "step": 18701
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4350486386301164e-05,
+      "loss": 2.0968,
+      "step": 18702
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4349946527367347e-05,
+      "loss": 2.0333,
+      "step": 18703
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4349406652796505e-05,
+      "loss": 2.0481,
+      "step": 18704
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4348866762590572e-05,
+      "loss": 1.6688,
+      "step": 18705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4348326856751496e-05,
+      "loss": 1.9634,
+      "step": 18706
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4347786935281209e-05,
+      "loss": 1.7353,
+      "step": 18707
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4347246998181656e-05,
+      "loss": 1.7578,
+      "step": 18708
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4346707045454782e-05,
+      "loss": 2.0285,
+      "step": 18709
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.434616707710252e-05,
+      "loss": 1.7877,
+      "step": 18710
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4345627093126816e-05,
+      "loss": 1.9121,
+      "step": 18711
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4345087093529612e-05,
+      "loss": 1.9121,
+      "step": 18712
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.434454707831285e-05,
+      "loss": 1.8372,
+      "step": 18713
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4344007047478463e-05,
+      "loss": 1.5197,
+      "step": 18714
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.43434670010284e-05,
+      "loss": 1.6049,
+      "step": 18715
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4342926938964603e-05,
+      "loss": 1.8764,
+      "step": 18716
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4342386861289007e-05,
+      "loss": 2.2006,
+      "step": 18717
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4341846768003558e-05,
+      "loss": 1.6367,
+      "step": 18718
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.43413066591102e-05,
+      "loss": 1.9485,
+      "step": 18719
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4340766534610867e-05,
+      "loss": 1.9255,
+      "step": 18720
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4340226394507506e-05,
+      "loss": 2.1963,
+      "step": 18721
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.433968623880206e-05,
+      "loss": 1.7431,
+      "step": 18722
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4339146067496466e-05,
+      "loss": 2.0521,
+      "step": 18723
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.4338605880592666e-05,
+      "loss": 1.3546,
+      "step": 18724
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4338065678092607e-05,
+      "loss": 1.9257,
+      "step": 18725
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.4337525459998228e-05,
+      "loss": 2.577,
+      "step": 18726
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.433698522631147e-05,
+      "loss": 1.4687,
+      "step": 18727
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4336444977034276e-05,
+      "loss": 2.0704,
+      "step": 18728
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4335904712168587e-05,
+      "loss": 2.0595,
+      "step": 18729
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4335364431716347e-05,
+      "loss": 2.3454,
+      "step": 18730
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4334824135679495e-05,
+      "loss": 1.8607,
+      "step": 18731
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4334283824059978e-05,
+      "loss": 1.8511,
+      "step": 18732
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4333743496859734e-05,
+      "loss": 1.8875,
+      "step": 18733
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4333203154080707e-05,
+      "loss": 2.1957,
+      "step": 18734
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4332662795724842e-05,
+      "loss": 1.9938,
+      "step": 18735
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4332122421794078e-05,
+      "loss": 2.2854,
+      "step": 18736
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4331582032290357e-05,
+      "loss": 1.8477,
+      "step": 18737
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.433104162721562e-05,
+      "loss": 1.6878,
+      "step": 18738
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4330501206571819e-05,
+      "loss": 1.5923,
+      "step": 18739
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4329960770360888e-05,
+      "loss": 1.9246,
+      "step": 18740
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.432942031858477e-05,
+      "loss": 2.0032,
+      "step": 18741
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4328879851245412e-05,
+      "loss": 2.0116,
+      "step": 18742
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.4328339368344755e-05,
+      "loss": 2.1172,
+      "step": 18743
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4327798869884739e-05,
+      "loss": 1.8869,
+      "step": 18744
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.432725835586731e-05,
+      "loss": 2.0581,
+      "step": 18745
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4326717826294414e-05,
+      "loss": 1.7068,
+      "step": 18746
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4326177281167986e-05,
+      "loss": 1.6474,
+      "step": 18747
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4325636720489976e-05,
+      "loss": 2.0884,
+      "step": 18748
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4325096144262327e-05,
+      "loss": 2.0178,
+      "step": 18749
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4324555552486978e-05,
+      "loss": 1.703,
+      "step": 18750
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4324014945165874e-05,
+      "loss": 1.7414,
+      "step": 18751
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.432347432230096e-05,
+      "loss": 1.9289,
+      "step": 18752
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4322933683894178e-05,
+      "loss": 2.0831,
+      "step": 18753
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.432239302994747e-05,
+      "loss": 1.8719,
+      "step": 18754
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4321852360462785e-05,
+      "loss": 2.2289,
+      "step": 18755
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.432131167544206e-05,
+      "loss": 1.92,
+      "step": 18756
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4320770974887244e-05,
+      "loss": 1.5727,
+      "step": 18757
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4320230258800276e-05,
+      "loss": 1.5185,
+      "step": 18758
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4319689527183102e-05,
+      "loss": 1.8586,
+      "step": 18759
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4319148780037667e-05,
+      "loss": 1.5165,
+      "step": 18760
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4318608017365909e-05,
+      "loss": 1.8472,
+      "step": 18761
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4318067239169782e-05,
+      "loss": 1.7883,
+      "step": 18762
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4317526445451223e-05,
+      "loss": 1.9984,
+      "step": 18763
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4316985636212177e-05,
+      "loss": 1.848,
+      "step": 18764
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4316444811454591e-05,
+      "loss": 1.8032,
+      "step": 18765
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.4315903971180402e-05,
+      "loss": 1.8036,
+      "step": 18766
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4315363115391563e-05,
+      "loss": 1.6311,
+      "step": 18767
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4314822244090012e-05,
+      "loss": 1.8101,
+      "step": 18768
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4314281357277697e-05,
+      "loss": 2.2546,
+      "step": 18769
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4313740454956557e-05,
+      "loss": 1.6629,
+      "step": 18770
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4313199537128543e-05,
+      "loss": 1.8015,
+      "step": 18771
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.43126586037956e-05,
+      "loss": 1.6848,
+      "step": 18772
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4312117654959666e-05,
+      "loss": 2.095,
+      "step": 18773
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4311576690622689e-05,
+      "loss": 1.6184,
+      "step": 18774
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4311035710786612e-05,
+      "loss": 1.6209,
+      "step": 18775
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4310494715453382e-05,
+      "loss": 2.0479,
+      "step": 18776
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4309953704624943e-05,
+      "loss": 1.8144,
+      "step": 18777
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.430941267830324e-05,
+      "loss": 1.7704,
+      "step": 18778
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4308871636490218e-05,
+      "loss": 1.8139,
+      "step": 18779
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4308330579187821e-05,
+      "loss": 1.8947,
+      "step": 18780
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4307789506397996e-05,
+      "loss": 1.6158,
+      "step": 18781
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.4307248418122682e-05,
+      "loss": 2.0339,
+      "step": 18782
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4306707314363834e-05,
+      "loss": 1.8652,
+      "step": 18783
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.430616619512339e-05,
+      "loss": 2.1161,
+      "step": 18784
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4305625060403294e-05,
+      "loss": 1.9459,
+      "step": 18785
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.43050839102055e-05,
+      "loss": 1.848,
+      "step": 18786
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4304542744531942e-05,
+      "loss": 1.825,
+      "step": 18787
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4304001563384573e-05,
+      "loss": 2.0332,
+      "step": 18788
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4303460366765334e-05,
+      "loss": 1.96,
+      "step": 18789
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4302919154676175e-05,
+      "loss": 1.9801,
+      "step": 18790
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.430237792711904e-05,
+      "loss": 1.6398,
+      "step": 18791
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4301836684095871e-05,
+      "loss": 1.7716,
+      "step": 18792
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.430129542560862e-05,
+      "loss": 1.7509,
+      "step": 18793
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4300754151659226e-05,
+      "loss": 2.0761,
+      "step": 18794
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.430021286224964e-05,
+      "loss": 1.9971,
+      "step": 18795
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4299671557381807e-05,
+      "loss": 1.768,
+      "step": 18796
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4299130237057668e-05,
+      "loss": 1.9376,
+      "step": 18797
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4298588901279176e-05,
+      "loss": 1.8582,
+      "step": 18798
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4298047550048269e-05,
+      "loss": 1.8054,
+      "step": 18799
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.4297506183366903e-05,
+      "loss": 1.6471,
+      "step": 18800
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4296964801237012e-05,
+      "loss": 1.8637,
+      "step": 18801
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4296423403660553e-05,
+      "loss": 1.7749,
+      "step": 18802
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4295881990639468e-05,
+      "loss": 1.7205,
+      "step": 18803
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4295340562175704e-05,
+      "loss": 1.8061,
+      "step": 18804
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4294799118271205e-05,
+      "loss": 1.9498,
+      "step": 18805
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.429425765892792e-05,
+      "loss": 1.9268,
+      "step": 18806
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4293716184147792e-05,
+      "loss": 1.6551,
+      "step": 18807
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.429317469393277e-05,
+      "loss": 1.8536,
+      "step": 18808
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4292633188284802e-05,
+      "loss": 1.6412,
+      "step": 18809
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4292091667205832e-05,
+      "loss": 1.8694,
+      "step": 18810
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.429155013069781e-05,
+      "loss": 1.6375,
+      "step": 18811
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4291008578762676e-05,
+      "loss": 1.8634,
+      "step": 18812
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4290467011402384e-05,
+      "loss": 1.9116,
+      "step": 18813
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4289925428618877e-05,
+      "loss": 1.911,
+      "step": 18814
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4289383830414099e-05,
+      "loss": 1.9382,
+      "step": 18815
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4288842216790007e-05,
+      "loss": 1.4965,
+      "step": 18816
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4288300587748535e-05,
+      "loss": 1.6664,
+      "step": 18817
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.428775894329164e-05,
+      "loss": 2.0371,
+      "step": 18818
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4287217283421267e-05,
+      "loss": 1.8575,
+      "step": 18819
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4286675608139363e-05,
+      "loss": 1.8839,
+      "step": 18820
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4286133917447869e-05,
+      "loss": 1.6659,
+      "step": 18821
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4285592211348738e-05,
+      "loss": 1.8127,
+      "step": 18822
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4285050489843922e-05,
+      "loss": 1.6524,
+      "step": 18823
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4284508752935356e-05,
+      "loss": 2.0562,
+      "step": 18824
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4283967000624996e-05,
+      "loss": 1.7649,
+      "step": 18825
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.428342523291479e-05,
+      "loss": 1.7484,
+      "step": 18826
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.428288344980668e-05,
+      "loss": 1.9454,
+      "step": 18827
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.428234165130262e-05,
+      "loss": 1.8112,
+      "step": 18828
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4281799837404553e-05,
+      "loss": 1.8566,
+      "step": 18829
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4281258008114432e-05,
+      "loss": 1.9967,
+      "step": 18830
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4280716163434195e-05,
+      "loss": 1.977,
+      "step": 18831
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4280174303365799e-05,
+      "loss": 1.8934,
+      "step": 18832
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4279632427911188e-05,
+      "loss": 1.9792,
+      "step": 18833
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.427909053707231e-05,
+      "loss": 1.9226,
+      "step": 18834
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4278548630851115e-05,
+      "loss": 1.7257,
+      "step": 18835
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4278006709249548e-05,
+      "loss": 1.6419,
+      "step": 18836
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.427746477226956e-05,
+      "loss": 1.977,
+      "step": 18837
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4276922819913094e-05,
+      "loss": 1.9195,
+      "step": 18838
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4276380852182106e-05,
+      "loss": 2.1229,
+      "step": 18839
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.427583886907854e-05,
+      "loss": 1.8457,
+      "step": 18840
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4275296870604342e-05,
+      "loss": 1.827,
+      "step": 18841
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4274754856761466e-05,
+      "loss": 1.8926,
+      "step": 18842
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4274212827551854e-05,
+      "loss": 1.7939,
+      "step": 18843
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.427367078297746e-05,
+      "loss": 1.9346,
+      "step": 18844
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4273128723040228e-05,
+      "loss": 1.7771,
+      "step": 18845
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.427258664774211e-05,
+      "loss": 1.9394,
+      "step": 18846
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4272044557085054e-05,
+      "loss": 1.8074,
+      "step": 18847
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4271502451071009e-05,
+      "loss": 2.0219,
+      "step": 18848
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4270960329701922e-05,
+      "loss": 1.8826,
+      "step": 18849
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4270418192979742e-05,
+      "loss": 1.6854,
+      "step": 18850
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.426987604090642e-05,
+      "loss": 2.2272,
+      "step": 18851
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.42693338734839e-05,
+      "loss": 2.0886,
+      "step": 18852
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4268791690714138e-05,
+      "loss": 2.4308,
+      "step": 18853
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.426824949259908e-05,
+      "loss": 1.9173,
+      "step": 18854
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.426770727914067e-05,
+      "loss": 1.8455,
+      "step": 18855
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4267165050340866e-05,
+      "loss": 2.0564,
+      "step": 18856
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.426662280620161e-05,
+      "loss": 2.0086,
+      "step": 18857
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4266080546724857e-05,
+      "loss": 1.8918,
+      "step": 18858
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.426553827191255e-05,
+      "loss": 2.0218,
+      "step": 18859
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4264995981766646e-05,
+      "loss": 2.0459,
+      "step": 18860
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4264453676289084e-05,
+      "loss": 1.9564,
+      "step": 18861
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4263911355481823e-05,
+      "loss": 1.8081,
+      "step": 18862
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4263369019346812e-05,
+      "loss": 2.0593,
+      "step": 18863
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4262826667885992e-05,
+      "loss": 1.9357,
+      "step": 18864
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.426228430110132e-05,
+      "loss": 1.7521,
+      "step": 18865
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4261741918994742e-05,
+      "loss": 1.9764,
+      "step": 18866
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4261199521568217e-05,
+      "loss": 1.829,
+      "step": 18867
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.426065710882368e-05,
+      "loss": 1.9692,
+      "step": 18868
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.4260114680763091e-05,
+      "loss": 1.9897,
+      "step": 18869
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.42595722373884e-05,
+      "loss": 2.0612,
+      "step": 18870
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.425902977870155e-05,
+      "loss": 1.6384,
+      "step": 18871
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4258487304704496e-05,
+      "loss": 1.9283,
+      "step": 18872
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4257944815399189e-05,
+      "loss": 1.766,
+      "step": 18873
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4257402310787577e-05,
+      "loss": 2.1177,
+      "step": 18874
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4256859790871609e-05,
+      "loss": 1.8294,
+      "step": 18875
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4256317255653237e-05,
+      "loss": 1.746,
+      "step": 18876
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4255774705134412e-05,
+      "loss": 1.7929,
+      "step": 18877
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4255232139317082e-05,
+      "loss": 1.6552,
+      "step": 18878
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4254689558203199e-05,
+      "loss": 1.9382,
+      "step": 18879
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4254146961794714e-05,
+      "loss": 1.4583,
+      "step": 18880
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4253604350093579e-05,
+      "loss": 1.7834,
+      "step": 18881
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.425306172310174e-05,
+      "loss": 1.6626,
+      "step": 18882
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4252519080821148e-05,
+      "loss": 1.891,
+      "step": 18883
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4251976423253759e-05,
+      "loss": 1.8086,
+      "step": 18884
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4251433750401518e-05,
+      "loss": 1.851,
+      "step": 18885
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.425089106226638e-05,
+      "loss": 1.8887,
+      "step": 18886
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.4250348358850292e-05,
+      "loss": 1.7069,
+      "step": 18887
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4249805640155208e-05,
+      "loss": 2.0321,
+      "step": 18888
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.424926290618308e-05,
+      "loss": 1.8676,
+      "step": 18889
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4248720156935853e-05,
+      "loss": 1.8858,
+      "step": 18890
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4248177392415483e-05,
+      "loss": 1.9086,
+      "step": 18891
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.424763461262392e-05,
+      "loss": 1.8278,
+      "step": 18892
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.4247091817563114e-05,
+      "loss": 1.9358,
+      "step": 18893
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.424654900723502e-05,
+      "loss": 1.6138,
+      "step": 18894
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4246006181641586e-05,
+      "loss": 2.5646,
+      "step": 18895
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4245463340784761e-05,
+      "loss": 1.7751,
+      "step": 18896
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4244920484666502e-05,
+      "loss": 1.8609,
+      "step": 18897
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4244377613288758e-05,
+      "loss": 1.7926,
+      "step": 18898
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4243834726653479e-05,
+      "loss": 1.9824,
+      "step": 18899
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.424329182476262e-05,
+      "loss": 1.7752,
+      "step": 18900
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4242748907618125e-05,
+      "loss": 1.7725,
+      "step": 18901
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4242205975221957e-05,
+      "loss": 1.8469,
+      "step": 18902
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4241663027576059e-05,
+      "loss": 2.4355,
+      "step": 18903
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4241120064682385e-05,
+      "loss": 1.7318,
+      "step": 18904
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.424057708654289e-05,
+      "loss": 1.8294,
+      "step": 18905
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.4240034093159518e-05,
+      "loss": 1.7877,
+      "step": 18906
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4239491084534234e-05,
+      "loss": 2.1477,
+      "step": 18907
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4238948060668975e-05,
+      "loss": 2.0009,
+      "step": 18908
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4238405021565704e-05,
+      "loss": 1.7595,
+      "step": 18909
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4237861967226368e-05,
+      "loss": 1.7,
+      "step": 18910
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4237318897652921e-05,
+      "loss": 1.8071,
+      "step": 18911
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4236775812847316e-05,
+      "loss": 1.7757,
+      "step": 18912
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.42362327128115e-05,
+      "loss": 1.9031,
+      "step": 18913
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.4235689597547433e-05,
+      "loss": 1.9774,
+      "step": 18914
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4235146467057062e-05,
+      "loss": 1.924,
+      "step": 18915
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4234603321342342e-05,
+      "loss": 1.8299,
+      "step": 18916
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4234060160405223e-05,
+      "loss": 2.1654,
+      "step": 18917
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.4233516984247659e-05,
+      "loss": 1.9635,
+      "step": 18918
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4232973792871604e-05,
+      "loss": 1.8671,
+      "step": 18919
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.423243058627901e-05,
+      "loss": 1.888,
+      "step": 18920
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4231887364471827e-05,
+      "loss": 1.8127,
+      "step": 18921
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.423134412745201e-05,
+      "loss": 1.8793,
+      "step": 18922
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4230800875221514e-05,
+      "loss": 2.0148,
+      "step": 18923
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4230257607782285e-05,
+      "loss": 1.8432,
+      "step": 18924
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4229714325136286e-05,
+      "loss": 1.7631,
+      "step": 18925
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.375,
+      "learning_rate": 1.422917102728546e-05,
+      "loss": 2.04,
+      "step": 18926
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4228627714231766e-05,
+      "loss": 1.8751,
+      "step": 18927
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4228084385977154e-05,
+      "loss": 1.7989,
+      "step": 18928
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4227541042523577e-05,
+      "loss": 1.7109,
+      "step": 18929
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4226997683872993e-05,
+      "loss": 1.6475,
+      "step": 18930
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4226454310027351e-05,
+      "loss": 1.6783,
+      "step": 18931
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4225910920988608e-05,
+      "loss": 1.8297,
+      "step": 18932
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.422536751675871e-05,
+      "loss": 2.0513,
+      "step": 18933
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4224824097339617e-05,
+      "loss": 1.8236,
+      "step": 18934
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4224280662733283e-05,
+      "loss": 1.7592,
+      "step": 18935
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.4223737212941656e-05,
+      "loss": 1.479,
+      "step": 18936
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4223193747966693e-05,
+      "loss": 1.825,
+      "step": 18937
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4222650267810349e-05,
+      "loss": 1.8737,
+      "step": 18938
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.4222106772474573e-05,
+      "loss": 2.5508,
+      "step": 18939
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4221563261961322e-05,
+      "loss": 1.9511,
+      "step": 18940
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4221019736272553e-05,
+      "loss": 1.6161,
+      "step": 18941
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4220476195410213e-05,
+      "loss": 1.6088,
+      "step": 18942
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4219932639376257e-05,
+      "loss": 2.4639,
+      "step": 18943
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4219389068172647e-05,
+      "loss": 1.6739,
+      "step": 18944
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4218845481801329e-05,
+      "loss": 1.9667,
+      "step": 18945
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4218301880264256e-05,
+      "loss": 1.9689,
+      "step": 18946
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.421775826356339e-05,
+      "loss": 1.9029,
+      "step": 18947
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.4217214631700677e-05,
+      "loss": 2.0083,
+      "step": 18948
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4216670984678077e-05,
+      "loss": 1.7737,
+      "step": 18949
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4216127322497539e-05,
+      "loss": 1.953,
+      "step": 18950
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4215583645161023e-05,
+      "loss": 1.9752,
+      "step": 18951
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4215039952670482e-05,
+      "loss": 1.7918,
+      "step": 18952
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4214496245027866e-05,
+      "loss": 1.7527,
+      "step": 18953
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4213952522235136e-05,
+      "loss": 1.6501,
+      "step": 18954
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.421340878429424e-05,
+      "loss": 2.2112,
+      "step": 18955
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4212865031207137e-05,
+      "loss": 1.4608,
+      "step": 18956
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4212321262975782e-05,
+      "loss": 1.8675,
+      "step": 18957
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4211777479602127e-05,
+      "loss": 1.7954,
+      "step": 18958
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4211233681088128e-05,
+      "loss": 1.7898,
+      "step": 18959
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4210689867435739e-05,
+      "loss": 1.8193,
+      "step": 18960
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4210146038646917e-05,
+      "loss": 1.9482,
+      "step": 18961
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4209602194723616e-05,
+      "loss": 1.7843,
+      "step": 18962
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.4209058335667789e-05,
+      "loss": 2.7332,
+      "step": 18963
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4208514461481393e-05,
+      "loss": 1.6869,
+      "step": 18964
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4207970572166385e-05,
+      "loss": 1.9757,
+      "step": 18965
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4207426667724717e-05,
+      "loss": 1.6932,
+      "step": 18966
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4206882748158341e-05,
+      "loss": 1.3971,
+      "step": 18967
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4206338813469223e-05,
+      "loss": 1.8562,
+      "step": 18968
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4205794863659309e-05,
+      "loss": 1.9443,
+      "step": 18969
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4205250898730555e-05,
+      "loss": 2.0346,
+      "step": 18970
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.420470691868492e-05,
+      "loss": 2.305,
+      "step": 18971
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.420416292352436e-05,
+      "loss": 1.6496,
+      "step": 18972
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4203618913250827e-05,
+      "loss": 1.8607,
+      "step": 18973
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4203074887866278e-05,
+      "loss": 2.0486,
+      "step": 18974
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4202530847372669e-05,
+      "loss": 1.5627,
+      "step": 18975
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4201986791771953e-05,
+      "loss": 2.1291,
+      "step": 18976
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.420144272106609e-05,
+      "loss": 1.7522,
+      "step": 18977
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4200898635257037e-05,
+      "loss": 1.8773,
+      "step": 18978
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4200354534346743e-05,
+      "loss": 1.9557,
+      "step": 18979
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.419981041833717e-05,
+      "loss": 2.1141,
+      "step": 18980
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4199266287230272e-05,
+      "loss": 1.6601,
+      "step": 18981
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4198722141028003e-05,
+      "loss": 1.759,
+      "step": 18982
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4198177979732322e-05,
+      "loss": 2.013,
+      "step": 18983
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.4197633803345183e-05,
+      "loss": 1.9887,
+      "step": 18984
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4197089611868544e-05,
+      "loss": 1.775,
+      "step": 18985
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.419654540530436e-05,
+      "loss": 1.751,
+      "step": 18986
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4196001183654589e-05,
+      "loss": 1.7996,
+      "step": 18987
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.4195456946921184e-05,
+      "loss": 2.1454,
+      "step": 18988
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4194912695106107e-05,
+      "loss": 2.0863,
+      "step": 18989
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4194368428211304e-05,
+      "loss": 1.7865,
+      "step": 18990
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4193824146238746e-05,
+      "loss": 1.7245,
+      "step": 18991
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.4193279849190378e-05,
+      "loss": 1.7527,
+      "step": 18992
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.419273553706816e-05,
+      "loss": 1.8662,
+      "step": 18993
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4192191209874052e-05,
+      "loss": 1.9431,
+      "step": 18994
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4191646867610007e-05,
+      "loss": 1.6934,
+      "step": 18995
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4191102510277983e-05,
+      "loss": 1.9179,
+      "step": 18996
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4190558137879934e-05,
+      "loss": 1.9998,
+      "step": 18997
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4190013750417827e-05,
+      "loss": 1.6223,
+      "step": 18998
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4189469347893603e-05,
+      "loss": 2.3241,
+      "step": 18999
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4188924930309232e-05,
+      "loss": 1.7323,
+      "step": 19000
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4188380497666664e-05,
+      "loss": 1.8417,
+      "step": 19001
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4187836049967861e-05,
+      "loss": 1.9004,
+      "step": 19002
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4187291587214777e-05,
+      "loss": 1.6132,
+      "step": 19003
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4186747109409367e-05,
+      "loss": 1.9829,
+      "step": 19004
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4186202616553597e-05,
+      "loss": 1.9699,
+      "step": 19005
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4185658108649413e-05,
+      "loss": 2.0313,
+      "step": 19006
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4185113585698781e-05,
+      "loss": 1.9309,
+      "step": 19007
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4184569047703657e-05,
+      "loss": 1.8531,
+      "step": 19008
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4184024494665995e-05,
+      "loss": 1.8332,
+      "step": 19009
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4183479926587755e-05,
+      "loss": 2.0548,
+      "step": 19010
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4182935343470895e-05,
+      "loss": 1.9341,
+      "step": 19011
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.418239074531737e-05,
+      "loss": 2.3534,
+      "step": 19012
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4181846132129136e-05,
+      "loss": 1.9003,
+      "step": 19013
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.4181301503908158e-05,
+      "loss": 1.6804,
+      "step": 19014
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4180756860656391e-05,
+      "loss": 2.0197,
+      "step": 19015
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.418021220237579e-05,
+      "loss": 1.9586,
+      "step": 19016
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4179667529068314e-05,
+      "loss": 1.8579,
+      "step": 19017
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4179122840735924e-05,
+      "loss": 1.5153,
+      "step": 19018
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4178578137380573e-05,
+      "loss": 1.825,
+      "step": 19019
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4178033419004225e-05,
+      "loss": 1.8654,
+      "step": 19020
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4177488685608833e-05,
+      "loss": 1.8848,
+      "step": 19021
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.417694393719636e-05,
+      "loss": 1.7782,
+      "step": 19022
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4176399173768756e-05,
+      "loss": 1.903,
+      "step": 19023
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.4175854395327986e-05,
+      "loss": 2.2198,
+      "step": 19024
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4175309601876009e-05,
+      "loss": 1.685,
+      "step": 19025
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4174764793414781e-05,
+      "loss": 1.6908,
+      "step": 19026
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4174219969946258e-05,
+      "loss": 1.8359,
+      "step": 19027
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4173675131472408e-05,
+      "loss": 1.8303,
+      "step": 19028
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.417313027799518e-05,
+      "loss": 1.6323,
+      "step": 19029
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4172585409516533e-05,
+      "loss": 1.8373,
+      "step": 19030
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4172040526038431e-05,
+      "loss": 1.8871,
+      "step": 19031
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4171495627562827e-05,
+      "loss": 1.826,
+      "step": 19032
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4170950714091685e-05,
+      "loss": 1.9818,
+      "step": 19033
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.417040578562696e-05,
+      "loss": 1.8621,
+      "step": 19034
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4169860842170616e-05,
+      "loss": 1.9099,
+      "step": 19035
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4169315883724606e-05,
+      "loss": 2.0355,
+      "step": 19036
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.416877091029089e-05,
+      "loss": 1.5983,
+      "step": 19037
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.4168225921871433e-05,
+      "loss": 2.25,
+      "step": 19038
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4167680918468183e-05,
+      "loss": 1.7703,
+      "step": 19039
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4167135900083108e-05,
+      "loss": 1.8706,
+      "step": 19040
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4166590866718168e-05,
+      "loss": 2.0057,
+      "step": 19041
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4166045818375318e-05,
+      "loss": 1.7386,
+      "step": 19042
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4165500755056518e-05,
+      "loss": 1.8413,
+      "step": 19043
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4164955676763725e-05,
+      "loss": 1.8995,
+      "step": 19044
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4164410583498907e-05,
+      "loss": 1.887,
+      "step": 19045
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4163865475264015e-05,
+      "loss": 1.8425,
+      "step": 19046
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.416332035206101e-05,
+      "loss": 1.809,
+      "step": 19047
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4162775213891854e-05,
+      "loss": 1.6929,
+      "step": 19048
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4162230060758507e-05,
+      "loss": 1.6694,
+      "step": 19049
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4161684892662925e-05,
+      "loss": 1.9233,
+      "step": 19050
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.416113970960707e-05,
+      "loss": 1.9705,
+      "step": 19051
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4160594511592904e-05,
+      "loss": 1.87,
+      "step": 19052
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4160049298622383e-05,
+      "loss": 1.754,
+      "step": 19053
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.415950407069747e-05,
+      "loss": 1.9433,
+      "step": 19054
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4158958827820122e-05,
+      "loss": 1.7211,
+      "step": 19055
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4158413569992303e-05,
+      "loss": 2.0285,
+      "step": 19056
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4157868297215968e-05,
+      "loss": 1.841,
+      "step": 19057
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.415732300949308e-05,
+      "loss": 1.6737,
+      "step": 19058
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.4156777706825601e-05,
+      "loss": 2.0643,
+      "step": 19059
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.4156232389215486e-05,
+      "loss": 2.08,
+      "step": 19060
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.41556870566647e-05,
+      "loss": 1.9732,
+      "step": 19061
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4155141709175203e-05,
+      "loss": 1.8654,
+      "step": 19062
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4154596346748952e-05,
+      "loss": 1.7842,
+      "step": 19063
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4154050969387914e-05,
+      "loss": 1.7371,
+      "step": 19064
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4153505577094043e-05,
+      "loss": 2.0107,
+      "step": 19065
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4152960169869299e-05,
+      "loss": 1.9706,
+      "step": 19066
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4152414747715646e-05,
+      "loss": 1.8236,
+      "step": 19067
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4151869310635048e-05,
+      "loss": 1.8999,
+      "step": 19068
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4151323858629458e-05,
+      "loss": 1.946,
+      "step": 19069
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4150778391700842e-05,
+      "loss": 1.8764,
+      "step": 19070
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4150232909851158e-05,
+      "loss": 1.8215,
+      "step": 19071
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.414968741308237e-05,
+      "loss": 1.6879,
+      "step": 19072
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4149141901396435e-05,
+      "loss": 1.7044,
+      "step": 19073
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4148596374795316e-05,
+      "loss": 1.8399,
+      "step": 19074
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4148050833280977e-05,
+      "loss": 1.9605,
+      "step": 19075
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4147505276855373e-05,
+      "loss": 1.8645,
+      "step": 19076
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.414695970552047e-05,
+      "loss": 1.7431,
+      "step": 19077
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4146414119278228e-05,
+      "loss": 1.9235,
+      "step": 19078
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4145868518130608e-05,
+      "loss": 1.7711,
+      "step": 19079
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.4145322902079572e-05,
+      "loss": 3.1754,
+      "step": 19080
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4144777271127077e-05,
+      "loss": 1.9421,
+      "step": 19081
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.414423162527509e-05,
+      "loss": 1.9475,
+      "step": 19082
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.375,
+      "learning_rate": 1.414368596452557e-05,
+      "loss": 1.9701,
+      "step": 19083
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.414314028888048e-05,
+      "loss": 1.9574,
+      "step": 19084
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.414259459834178e-05,
+      "loss": 1.9418,
+      "step": 19085
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4142048892911432e-05,
+      "loss": 1.8531,
+      "step": 19086
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4141503172591398e-05,
+      "loss": 1.9971,
+      "step": 19087
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4140957437383638e-05,
+      "loss": 1.9438,
+      "step": 19088
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4140411687290119e-05,
+      "loss": 1.9668,
+      "step": 19089
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4139865922312795e-05,
+      "loss": 2.0386,
+      "step": 19090
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4139320142453633e-05,
+      "loss": 1.9411,
+      "step": 19091
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.4138774347714597e-05,
+      "loss": 1.7034,
+      "step": 19092
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4138228538097642e-05,
+      "loss": 1.9776,
+      "step": 19093
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4137682713604737e-05,
+      "loss": 1.7996,
+      "step": 19094
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.413713687423784e-05,
+      "loss": 1.7941,
+      "step": 19095
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4136591019998915e-05,
+      "loss": 1.889,
+      "step": 19096
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.413604515088992e-05,
+      "loss": 2.1091,
+      "step": 19097
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4135499266912824e-05,
+      "loss": 1.6708,
+      "step": 19098
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4134953368069591e-05,
+      "loss": 1.7357,
+      "step": 19099
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4134407454362168e-05,
+      "loss": 2.1287,
+      "step": 19100
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4133861525792536e-05,
+      "loss": 1.6567,
+      "step": 19101
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4133315582362648e-05,
+      "loss": 1.7722,
+      "step": 19102
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4132769624074465e-05,
+      "loss": 2.0561,
+      "step": 19103
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4132223650929952e-05,
+      "loss": 1.7702,
+      "step": 19104
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4131677662931077e-05,
+      "loss": 1.9112,
+      "step": 19105
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4131131660079798e-05,
+      "loss": 1.6352,
+      "step": 19106
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4130585642378072e-05,
+      "loss": 1.9198,
+      "step": 19107
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.4130039609827872e-05,
+      "loss": 2.0695,
+      "step": 19108
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4129493562431156e-05,
+      "loss": 1.987,
+      "step": 19109
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4128947500189884e-05,
+      "loss": 1.6135,
+      "step": 19110
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4128401423106025e-05,
+      "loss": 1.9606,
+      "step": 19111
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4127855331181542e-05,
+      "loss": 1.7165,
+      "step": 19112
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.412730922441839e-05,
+      "loss": 1.7883,
+      "step": 19113
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.412676310281854e-05,
+      "loss": 1.9022,
+      "step": 19114
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4126216966383953e-05,
+      "loss": 1.4985,
+      "step": 19115
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.412567081511659e-05,
+      "loss": 1.6912,
+      "step": 19116
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4125124649018415e-05,
+      "loss": 2.0108,
+      "step": 19117
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4124578468091396e-05,
+      "loss": 2.4482,
+      "step": 19118
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.4124032272337491e-05,
+      "loss": 1.9525,
+      "step": 19119
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4123486061758666e-05,
+      "loss": 1.5022,
+      "step": 19120
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4122939836356881e-05,
+      "loss": 1.8053,
+      "step": 19121
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4122393596134107e-05,
+      "loss": 1.9875,
+      "step": 19122
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4121847341092298e-05,
+      "loss": 2.0299,
+      "step": 19123
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4121301071233427e-05,
+      "loss": 1.8267,
+      "step": 19124
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4120754786559451e-05,
+      "loss": 1.7687,
+      "step": 19125
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4120208487072338e-05,
+      "loss": 2.2144,
+      "step": 19126
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4119662172774048e-05,
+      "loss": 1.7738,
+      "step": 19127
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4119115843666546e-05,
+      "loss": 2.0159,
+      "step": 19128
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.41185694997518e-05,
+      "loss": 1.9228,
+      "step": 19129
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4118023141031767e-05,
+      "loss": 1.7928,
+      "step": 19130
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4117476767508416e-05,
+      "loss": 1.8776,
+      "step": 19131
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.411693037918371e-05,
+      "loss": 1.6887,
+      "step": 19132
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4116383976059614e-05,
+      "loss": 1.9342,
+      "step": 19133
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.411583755813809e-05,
+      "loss": 1.9903,
+      "step": 19134
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4115291125421103e-05,
+      "loss": 2.0048,
+      "step": 19135
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.411474467791062e-05,
+      "loss": 1.7688,
+      "step": 19136
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.4114198215608599e-05,
+      "loss": 1.6454,
+      "step": 19137
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4113651738517011e-05,
+      "loss": 2.1565,
+      "step": 19138
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4113105246637818e-05,
+      "loss": 1.8509,
+      "step": 19139
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4112558739972983e-05,
+      "loss": 1.8745,
+      "step": 19140
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4112012218524474e-05,
+      "loss": 1.6838,
+      "step": 19141
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.411146568229425e-05,
+      "loss": 1.8808,
+      "step": 19142
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4110919131284282e-05,
+      "loss": 1.8515,
+      "step": 19143
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.411037256549653e-05,
+      "loss": 1.899,
+      "step": 19144
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.410982598493296e-05,
+      "loss": 1.863,
+      "step": 19145
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4109279389595543e-05,
+      "loss": 2.0045,
+      "step": 19146
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4108732779486234e-05,
+      "loss": 1.9021,
+      "step": 19147
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4108186154607004e-05,
+      "loss": 1.932,
+      "step": 19148
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4107639514959815e-05,
+      "loss": 2.0755,
+      "step": 19149
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4107092860546633e-05,
+      "loss": 1.9915,
+      "step": 19150
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.4106546191369424e-05,
+      "loss": 1.87,
+      "step": 19151
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4105999507430155e-05,
+      "loss": 1.8332,
+      "step": 19152
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4105452808730786e-05,
+      "loss": 1.5699,
+      "step": 19153
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4104906095273286e-05,
+      "loss": 1.8761,
+      "step": 19154
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.410435936705962e-05,
+      "loss": 2.0507,
+      "step": 19155
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4103812624091751e-05,
+      "loss": 1.7485,
+      "step": 19156
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.4103265866371646e-05,
+      "loss": 1.3259,
+      "step": 19157
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4102719093901271e-05,
+      "loss": 1.5638,
+      "step": 19158
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4102172306682595e-05,
+      "loss": 1.8816,
+      "step": 19159
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4101625504717576e-05,
+      "loss": 1.749,
+      "step": 19160
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4101078688008183e-05,
+      "loss": 2.3605,
+      "step": 19161
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4100531856556382e-05,
+      "loss": 1.8317,
+      "step": 19162
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.409998501036414e-05,
+      "loss": 1.8513,
+      "step": 19163
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4099438149433422e-05,
+      "loss": 1.9316,
+      "step": 19164
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.409889127376619e-05,
+      "loss": 2.123,
+      "step": 19165
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.4098344383364419e-05,
+      "loss": 1.5832,
+      "step": 19166
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4097797478230065e-05,
+      "loss": 1.9498,
+      "step": 19167
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.40972505583651e-05,
+      "loss": 1.6976,
+      "step": 19168
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4096703623771486e-05,
+      "loss": 1.8954,
+      "step": 19169
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4096156674451194e-05,
+      "loss": 2.0505,
+      "step": 19170
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4095609710406186e-05,
+      "loss": 1.922,
+      "step": 19171
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4095062731638428e-05,
+      "loss": 1.6479,
+      "step": 19172
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.4094515738149892e-05,
+      "loss": 2.098,
+      "step": 19173
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4093968729942535e-05,
+      "loss": 1.9671,
+      "step": 19174
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.409342170701833e-05,
+      "loss": 2.1208,
+      "step": 19175
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4092874669379246e-05,
+      "loss": 1.7555,
+      "step": 19176
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.4092327617027243e-05,
+      "loss": 1.7763,
+      "step": 19177
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4091780549964291e-05,
+      "loss": 1.676,
+      "step": 19178
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4091233468192354e-05,
+      "loss": 1.896,
+      "step": 19179
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4090686371713403e-05,
+      "loss": 1.8748,
+      "step": 19180
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4090139260529397e-05,
+      "loss": 1.9869,
+      "step": 19181
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.4089592134642309e-05,
+      "loss": 1.7962,
+      "step": 19182
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4089044994054108e-05,
+      "loss": 1.8708,
+      "step": 19183
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.4088497838766755e-05,
+      "loss": 2.1063,
+      "step": 19184
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4087950668782217e-05,
+      "loss": 1.8279,
+      "step": 19185
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4087403484102468e-05,
+      "loss": 1.9046,
+      "step": 19186
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4086856284729468e-05,
+      "loss": 1.5668,
+      "step": 19187
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4086309070665183e-05,
+      "loss": 1.9088,
+      "step": 19188
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.408576184191159e-05,
+      "loss": 1.9542,
+      "step": 19189
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.4085214598470645e-05,
+      "loss": 1.6664,
+      "step": 19190
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4084667340344319e-05,
+      "loss": 2.0789,
+      "step": 19191
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4084120067534582e-05,
+      "loss": 1.7803,
+      "step": 19192
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.4083572780043397e-05,
+      "loss": 2.168,
+      "step": 19193
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4083025477872733e-05,
+      "loss": 1.5902,
+      "step": 19194
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4082478161024559e-05,
+      "loss": 1.6819,
+      "step": 19195
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4081930829500844e-05,
+      "loss": 1.8647,
+      "step": 19196
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4081383483303553e-05,
+      "loss": 2.0303,
+      "step": 19197
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.408083612243465e-05,
+      "loss": 1.8682,
+      "step": 19198
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.408028874689611e-05,
+      "loss": 1.9247,
+      "step": 19199
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.4079741356689893e-05,
+      "loss": 1.7649,
+      "step": 19200
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4079193951817973e-05,
+      "loss": 1.9242,
+      "step": 19201
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4078646532282316e-05,
+      "loss": 1.6883,
+      "step": 19202
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.407809909808489e-05,
+      "loss": 1.9358,
+      "step": 19203
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.407755164922766e-05,
+      "loss": 2.0437,
+      "step": 19204
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4077004185712596e-05,
+      "loss": 2.2167,
+      "step": 19205
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4076456707541669e-05,
+      "loss": 1.9068,
+      "step": 19206
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4075909214716838e-05,
+      "loss": 1.9307,
+      "step": 19207
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.4075361707240082e-05,
+      "loss": 1.5378,
+      "step": 19208
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4074814185113366e-05,
+      "loss": 1.8829,
+      "step": 19209
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4074266648338654e-05,
+      "loss": 2.0195,
+      "step": 19210
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4073719096917917e-05,
+      "loss": 1.9164,
+      "step": 19211
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4073171530853122e-05,
+      "loss": 1.7842,
+      "step": 19212
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4072623950146243e-05,
+      "loss": 1.8368,
+      "step": 19213
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.407207635479924e-05,
+      "loss": 1.8099,
+      "step": 19214
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4071528744814088e-05,
+      "loss": 1.7709,
+      "step": 19215
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4070981120192753e-05,
+      "loss": 1.8831,
+      "step": 19216
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.40704334809372e-05,
+      "loss": 1.7048,
+      "step": 19217
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.4069885827049405e-05,
+      "loss": 2.1916,
+      "step": 19218
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4069338158531332e-05,
+      "loss": 1.9033,
+      "step": 19219
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.406879047538495e-05,
+      "loss": 2.2076,
+      "step": 19220
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.406824277761223e-05,
+      "loss": 2.189,
+      "step": 19221
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4067695065215137e-05,
+      "loss": 1.8882,
+      "step": 19222
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4067147338195644e-05,
+      "loss": 1.8023,
+      "step": 19223
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4066599596555719e-05,
+      "loss": 1.8997,
+      "step": 19224
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4066051840297328e-05,
+      "loss": 1.9198,
+      "step": 19225
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4065504069422441e-05,
+      "loss": 1.8435,
+      "step": 19226
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4064956283933033e-05,
+      "loss": 1.9319,
+      "step": 19227
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4064408483831065e-05,
+      "loss": 1.8493,
+      "step": 19228
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.406386066911851e-05,
+      "loss": 1.9605,
+      "step": 19229
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.4063312839797339e-05,
+      "loss": 1.9353,
+      "step": 19230
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4062764995869518e-05,
+      "loss": 1.7483,
+      "step": 19231
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4062217137337018e-05,
+      "loss": 1.8803,
+      "step": 19232
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.4061669264201807e-05,
+      "loss": 1.6434,
+      "step": 19233
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4061121376465856e-05,
+      "loss": 1.7798,
+      "step": 19234
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.4060573474131132e-05,
+      "loss": 2.1447,
+      "step": 19235
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4060025557199612e-05,
+      "loss": 2.1018,
+      "step": 19236
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4059477625673257e-05,
+      "loss": 2.0237,
+      "step": 19237
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4058929679554038e-05,
+      "loss": 1.6462,
+      "step": 19238
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.405838171884393e-05,
+      "loss": 1.2775,
+      "step": 19239
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.40578337435449e-05,
+      "loss": 1.7494,
+      "step": 19240
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4057285753658916e-05,
+      "loss": 1.7901,
+      "step": 19241
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4056737749187948e-05,
+      "loss": 1.8582,
+      "step": 19242
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4056189730133968e-05,
+      "loss": 2.1155,
+      "step": 19243
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4055641696498945e-05,
+      "loss": 1.8652,
+      "step": 19244
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.405509364828485e-05,
+      "loss": 2.0058,
+      "step": 19245
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4054545585493651e-05,
+      "loss": 1.64,
+      "step": 19246
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4053997508127321e-05,
+      "loss": 2.0554,
+      "step": 19247
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.405344941618783e-05,
+      "loss": 1.9479,
+      "step": 19248
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4052901309677143e-05,
+      "loss": 1.6446,
+      "step": 19249
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4052353188597238e-05,
+      "loss": 1.7574,
+      "step": 19250
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4051805052950078e-05,
+      "loss": 1.9763,
+      "step": 19251
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.405125690273764e-05,
+      "loss": 1.7516,
+      "step": 19252
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4050708737961891e-05,
+      "loss": 1.8167,
+      "step": 19253
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4050160558624802e-05,
+      "loss": 1.8527,
+      "step": 19254
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.4049612364728343e-05,
+      "loss": 1.8659,
+      "step": 19255
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4049064156274483e-05,
+      "loss": 1.9102,
+      "step": 19256
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.40485159332652e-05,
+      "loss": 1.9267,
+      "step": 19257
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4047967695702456e-05,
+      "loss": 1.9592,
+      "step": 19258
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4047419443588226e-05,
+      "loss": 1.8635,
+      "step": 19259
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.4046871176924483e-05,
+      "loss": 1.7012,
+      "step": 19260
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4046322895713191e-05,
+      "loss": 1.692,
+      "step": 19261
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4045774599956327e-05,
+      "loss": 1.9894,
+      "step": 19262
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.4045226289655862e-05,
+      "loss": 1.55,
+      "step": 19263
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4044677964813765e-05,
+      "loss": 1.8581,
+      "step": 19264
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4044129625432002e-05,
+      "loss": 1.6102,
+      "step": 19265
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4043581271512554e-05,
+      "loss": 1.9498,
+      "step": 19266
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.4043032903057387e-05,
+      "loss": 1.9743,
+      "step": 19267
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.375,
+      "learning_rate": 1.4042484520068471e-05,
+      "loss": 1.9919,
+      "step": 19268
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4041936122547784e-05,
+      "loss": 2.0986,
+      "step": 19269
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4041387710497288e-05,
+      "loss": 1.7457,
+      "step": 19270
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.4040839283918962e-05,
+      "loss": 1.9336,
+      "step": 19271
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.404029084281477e-05,
+      "loss": 1.8678,
+      "step": 19272
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4039742387186692e-05,
+      "loss": 2.0501,
+      "step": 19273
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4039193917036694e-05,
+      "loss": 1.7018,
+      "step": 19274
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.403864543236675e-05,
+      "loss": 1.9475,
+      "step": 19275
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.4038096933178828e-05,
+      "loss": 1.718,
+      "step": 19276
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4037548419474905e-05,
+      "loss": 2.0378,
+      "step": 19277
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4036999891256948e-05,
+      "loss": 1.7063,
+      "step": 19278
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4036451348526935e-05,
+      "loss": 1.7969,
+      "step": 19279
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4035902791286834e-05,
+      "loss": 1.8038,
+      "step": 19280
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4035354219538613e-05,
+      "loss": 1.8145,
+      "step": 19281
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4034805633284248e-05,
+      "loss": 1.5837,
+      "step": 19282
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4034257032525716e-05,
+      "loss": 1.777,
+      "step": 19283
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4033708417264977e-05,
+      "loss": 1.8443,
+      "step": 19284
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4033159787504013e-05,
+      "loss": 2.1508,
+      "step": 19285
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4032611143244796e-05,
+      "loss": 1.9895,
+      "step": 19286
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4032062484489295e-05,
+      "loss": 2.0191,
+      "step": 19287
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4031513811239481e-05,
+      "loss": 1.6618,
+      "step": 19288
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.4030965123497327e-05,
+      "loss": 1.8902,
+      "step": 19289
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4030416421264812e-05,
+      "loss": 1.6393,
+      "step": 19290
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4029867704543898e-05,
+      "loss": 1.6559,
+      "step": 19291
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4029318973336565e-05,
+      "loss": 1.8756,
+      "step": 19292
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.4028770227644782e-05,
+      "loss": 1.5617,
+      "step": 19293
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4028221467470524e-05,
+      "loss": 1.6588,
+      "step": 19294
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4027672692815762e-05,
+      "loss": 1.8596,
+      "step": 19295
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.402712390368247e-05,
+      "loss": 1.6372,
+      "step": 19296
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.402657510007262e-05,
+      "loss": 1.9849,
+      "step": 19297
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.4026026281988182e-05,
+      "loss": 1.9051,
+      "step": 19298
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4025477449431133e-05,
+      "loss": 1.7168,
+      "step": 19299
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.4024928602403446e-05,
+      "loss": 1.7029,
+      "step": 19300
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.402437974090709e-05,
+      "loss": 1.8372,
+      "step": 19301
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.4023830864944043e-05,
+      "loss": 2.0433,
+      "step": 19302
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4023281974516274e-05,
+      "loss": 1.7592,
+      "step": 19303
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.402273306962576e-05,
+      "loss": 1.7922,
+      "step": 19304
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.4022184150274469e-05,
+      "loss": 1.8694,
+      "step": 19305
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.402163521646438e-05,
+      "loss": 1.5162,
+      "step": 19306
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.4021086268197462e-05,
+      "loss": 1.7786,
+      "step": 19307
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4020537305475692e-05,
+      "loss": 1.8935,
+      "step": 19308
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.401998832830104e-05,
+      "loss": 1.7613,
+      "step": 19309
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.401943933667548e-05,
+      "loss": 1.6639,
+      "step": 19310
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4018890330600984e-05,
+      "loss": 1.862,
+      "step": 19311
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.401834131007953e-05,
+      "loss": 1.8497,
+      "step": 19312
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4017792275113089e-05,
+      "loss": 1.8503,
+      "step": 19313
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4017243225703636e-05,
+      "loss": 1.9394,
+      "step": 19314
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4016694161853142e-05,
+      "loss": 1.9154,
+      "step": 19315
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4016145083563584e-05,
+      "loss": 1.8054,
+      "step": 19316
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4015595990836932e-05,
+      "loss": 1.6559,
+      "step": 19317
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.4015046883675165e-05,
+      "loss": 1.865,
+      "step": 19318
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.401449776208025e-05,
+      "loss": 1.7438,
+      "step": 19319
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.401394862605417e-05,
+      "loss": 1.4986,
+      "step": 19320
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.401339947559889e-05,
+      "loss": 1.6442,
+      "step": 19321
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.401285031071639e-05,
+      "loss": 1.8784,
+      "step": 19322
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4012301131408641e-05,
+      "loss": 1.7632,
+      "step": 19323
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.401175193767762e-05,
+      "loss": 2.0532,
+      "step": 19324
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.4011202729525297e-05,
+      "loss": 1.9399,
+      "step": 19325
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4010653506953648e-05,
+      "loss": 2.0914,
+      "step": 19326
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.4010104269964655e-05,
+      "loss": 1.6442,
+      "step": 19327
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.400955501856028e-05,
+      "loss": 2.1672,
+      "step": 19328
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4009005752742501e-05,
+      "loss": 1.9349,
+      "step": 19329
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.4008456472513299e-05,
+      "loss": 1.7019,
+      "step": 19330
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4007907177874643e-05,
+      "loss": 1.859,
+      "step": 19331
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.4007357868828506e-05,
+      "loss": 1.9545,
+      "step": 19332
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4006808545376867e-05,
+      "loss": 1.4946,
+      "step": 19333
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.40062592075217e-05,
+      "loss": 1.8444,
+      "step": 19334
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.4005709855264976e-05,
+      "loss": 1.7224,
+      "step": 19335
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.4005160488608674e-05,
+      "loss": 2.0234,
+      "step": 19336
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.4004611107554768e-05,
+      "loss": 1.8082,
+      "step": 19337
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.400406171210523e-05,
+      "loss": 2.071,
+      "step": 19338
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4003512302262039e-05,
+      "loss": 1.8377,
+      "step": 19339
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.4002962878027164e-05,
+      "loss": 1.8848,
+      "step": 19340
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.400241343940259e-05,
+      "loss": 1.7662,
+      "step": 19341
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.4001863986390283e-05,
+      "loss": 2.2169,
+      "step": 19342
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4001314518992222e-05,
+      "loss": 1.786,
+      "step": 19343
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4000765037210383e-05,
+      "loss": 1.5907,
+      "step": 19344
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4000215541046738e-05,
+      "loss": 1.7638,
+      "step": 19345
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3999666030503265e-05,
+      "loss": 1.4989,
+      "step": 19346
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3999116505581938e-05,
+      "loss": 1.8943,
+      "step": 19347
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3998566966284734e-05,
+      "loss": 1.6646,
+      "step": 19348
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3998017412613625e-05,
+      "loss": 1.8696,
+      "step": 19349
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.399746784457059e-05,
+      "loss": 2.0132,
+      "step": 19350
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3996918262157606e-05,
+      "loss": 1.7961,
+      "step": 19351
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3996368665376641e-05,
+      "loss": 1.9224,
+      "step": 19352
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.399581905422968e-05,
+      "loss": 2.0284,
+      "step": 19353
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3995269428718694e-05,
+      "loss": 3.3447,
+      "step": 19354
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3994719788845657e-05,
+      "loss": 1.7525,
+      "step": 19355
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3994170134612548e-05,
+      "loss": 2.0128,
+      "step": 19356
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3993620466021345e-05,
+      "loss": 1.8736,
+      "step": 19357
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.3993070783074019e-05,
+      "loss": 1.9129,
+      "step": 19358
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3992521085772543e-05,
+      "loss": 1.972,
+      "step": 19359
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3991971374118905e-05,
+      "loss": 2.099,
+      "step": 19360
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.3991421648115071e-05,
+      "loss": 1.5967,
+      "step": 19361
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.399087190776302e-05,
+      "loss": 2.0016,
+      "step": 19362
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3990322153064728e-05,
+      "loss": 1.8166,
+      "step": 19363
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3989772384022173e-05,
+      "loss": 1.9111,
+      "step": 19364
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.398922260063733e-05,
+      "loss": 2.0505,
+      "step": 19365
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.398867280291217e-05,
+      "loss": 1.6418,
+      "step": 19366
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3988122990848682e-05,
+      "loss": 2.0629,
+      "step": 19367
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.398757316444883e-05,
+      "loss": 2.1543,
+      "step": 19368
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3987023323714598e-05,
+      "loss": 1.6547,
+      "step": 19369
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3986473468647958e-05,
+      "loss": 1.9904,
+      "step": 19370
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.398592359925089e-05,
+      "loss": 1.5966,
+      "step": 19371
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.398537371552537e-05,
+      "loss": 2.3083,
+      "step": 19372
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3984823817473373e-05,
+      "loss": 1.9361,
+      "step": 19373
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3984273905096877e-05,
+      "loss": 1.7408,
+      "step": 19374
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3983723978397858e-05,
+      "loss": 1.7225,
+      "step": 19375
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3983174037378292e-05,
+      "loss": 1.9759,
+      "step": 19376
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3982624082040162e-05,
+      "loss": 2.0809,
+      "step": 19377
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.3982074112385438e-05,
+      "loss": 2.1985,
+      "step": 19378
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3981524128416097e-05,
+      "loss": 2.1574,
+      "step": 19379
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.398097413013412e-05,
+      "loss": 1.5236,
+      "step": 19380
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3980424117541485e-05,
+      "loss": 1.7952,
+      "step": 19381
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3979874090640163e-05,
+      "loss": 1.876,
+      "step": 19382
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3979324049432135e-05,
+      "loss": 1.8956,
+      "step": 19383
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.397877399391938e-05,
+      "loss": 2.031,
+      "step": 19384
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3978223924103873e-05,
+      "loss": 1.8528,
+      "step": 19385
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3977673839987591e-05,
+      "loss": 1.6407,
+      "step": 19386
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3977123741572512e-05,
+      "loss": 2.0616,
+      "step": 19387
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3976573628860617e-05,
+      "loss": 1.9046,
+      "step": 19388
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3976023501853875e-05,
+      "loss": 1.9451,
+      "step": 19389
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3975473360554273e-05,
+      "loss": 1.7601,
+      "step": 19390
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3974923204963781e-05,
+      "loss": 1.7594,
+      "step": 19391
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3974373035084385e-05,
+      "loss": 1.7507,
+      "step": 19392
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3973822850918055e-05,
+      "loss": 1.6232,
+      "step": 19393
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3973272652466771e-05,
+      "loss": 1.7161,
+      "step": 19394
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3972722439732514e-05,
+      "loss": 2.0679,
+      "step": 19395
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3972172212717254e-05,
+      "loss": 1.8696,
+      "step": 19396
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3971621971422979e-05,
+      "loss": 1.5573,
+      "step": 19397
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.397107171585166e-05,
+      "loss": 1.9698,
+      "step": 19398
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3970521446005278e-05,
+      "loss": 1.7026,
+      "step": 19399
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3969971161885812e-05,
+      "loss": 1.6057,
+      "step": 19400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5,
+      "learning_rate": 1.3969420863495238e-05,
+      "loss": 2.166,
+      "step": 19401
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3968870550835533e-05,
+      "loss": 1.7916,
+      "step": 19402
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.3968320223908678e-05,
+      "loss": 1.8465,
+      "step": 19403
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.396776988271665e-05,
+      "loss": 2.217,
+      "step": 19404
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.396721952726143e-05,
+      "loss": 1.8643,
+      "step": 19405
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.396666915754499e-05,
+      "loss": 2.0946,
+      "step": 19406
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3966118773569315e-05,
+      "loss": 1.7369,
+      "step": 19407
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.396556837533638e-05,
+      "loss": 1.8342,
+      "step": 19408
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3965017962848166e-05,
+      "loss": 1.9855,
+      "step": 19409
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3964467536106647e-05,
+      "loss": 1.9325,
+      "step": 19410
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3963917095113807e-05,
+      "loss": 1.8761,
+      "step": 19411
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.396336663987162e-05,
+      "loss": 1.717,
+      "step": 19412
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3962816170382072e-05,
+      "loss": 2.0778,
+      "step": 19413
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3962265686647134e-05,
+      "loss": 2.0983,
+      "step": 19414
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.396171518866879e-05,
+      "loss": 1.9846,
+      "step": 19415
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3961164676449015e-05,
+      "loss": 1.696,
+      "step": 19416
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3960614149989789e-05,
+      "loss": 1.9824,
+      "step": 19417
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3960063609293095e-05,
+      "loss": 1.8421,
+      "step": 19418
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3959513054360906e-05,
+      "loss": 2.2815,
+      "step": 19419
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3958962485195203e-05,
+      "loss": 2.1023,
+      "step": 19420
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3958411901797968e-05,
+      "loss": 1.7527,
+      "step": 19421
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3957861304171179e-05,
+      "loss": 1.7493,
+      "step": 19422
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3957310692316813e-05,
+      "loss": 2.0574,
+      "step": 19423
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.395676006623685e-05,
+      "loss": 2.0535,
+      "step": 19424
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3956209425933274e-05,
+      "loss": 1.591,
+      "step": 19425
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3955658771408058e-05,
+      "loss": 1.9168,
+      "step": 19426
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3955108102663185e-05,
+      "loss": 1.9577,
+      "step": 19427
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3954557419700633e-05,
+      "loss": 1.5272,
+      "step": 19428
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3954006722522384e-05,
+      "loss": 2.0443,
+      "step": 19429
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3953456011130415e-05,
+      "loss": 1.7369,
+      "step": 19430
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3952905285526706e-05,
+      "loss": 2.0361,
+      "step": 19431
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3952354545713238e-05,
+      "loss": 1.7919,
+      "step": 19432
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.395180379169199e-05,
+      "loss": 1.6073,
+      "step": 19433
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3951253023464942e-05,
+      "loss": 1.7946,
+      "step": 19434
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3950702241034075e-05,
+      "loss": 2.3073,
+      "step": 19435
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3950151444401364e-05,
+      "loss": 1.7179,
+      "step": 19436
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3949600633568796e-05,
+      "loss": 1.8789,
+      "step": 19437
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3949049808538346e-05,
+      "loss": 1.8569,
+      "step": 19438
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3948498969311999e-05,
+      "loss": 1.7515,
+      "step": 19439
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3947948115891728e-05,
+      "loss": 1.7677,
+      "step": 19440
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3947397248279519e-05,
+      "loss": 1.8827,
+      "step": 19441
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3946846366477354e-05,
+      "loss": 2.1636,
+      "step": 19442
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3946295470487204e-05,
+      "loss": 1.6761,
+      "step": 19443
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3945744560311056e-05,
+      "loss": 2.099,
+      "step": 19444
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.394519363595089e-05,
+      "loss": 1.6878,
+      "step": 19445
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3944642697408687e-05,
+      "loss": 1.962,
+      "step": 19446
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3944091744686428e-05,
+      "loss": 2.0527,
+      "step": 19447
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3943540777786089e-05,
+      "loss": 2.0856,
+      "step": 19448
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3942989796709654e-05,
+      "loss": 2.0113,
+      "step": 19449
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3942438801459105e-05,
+      "loss": 1.8656,
+      "step": 19450
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.394188779203642e-05,
+      "loss": 1.4758,
+      "step": 19451
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3941336768443577e-05,
+      "loss": 1.8332,
+      "step": 19452
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3940785730682565e-05,
+      "loss": 1.9436,
+      "step": 19453
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3940234678755358e-05,
+      "loss": 1.7516,
+      "step": 19454
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3939683612663942e-05,
+      "loss": 1.8136,
+      "step": 19455
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3939132532410292e-05,
+      "loss": 1.8668,
+      "step": 19456
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3938581437996392e-05,
+      "loss": 1.7129,
+      "step": 19457
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3938030329424225e-05,
+      "loss": 1.7623,
+      "step": 19458
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3937479206695769e-05,
+      "loss": 1.8873,
+      "step": 19459
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3936928069813006e-05,
+      "loss": 1.8398,
+      "step": 19460
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.393637691877792e-05,
+      "loss": 1.9247,
+      "step": 19461
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3935825753592489e-05,
+      "loss": 2.0153,
+      "step": 19462
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.3935274574258695e-05,
+      "loss": 2.2617,
+      "step": 19463
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3934723380778517e-05,
+      "loss": 1.5729,
+      "step": 19464
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3934172173153944e-05,
+      "loss": 1.868,
+      "step": 19465
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3933620951386949e-05,
+      "loss": 1.9755,
+      "step": 19466
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3933069715479517e-05,
+      "loss": 1.8644,
+      "step": 19467
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.393251846543363e-05,
+      "loss": 1.8342,
+      "step": 19468
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.393196720125127e-05,
+      "loss": 1.6019,
+      "step": 19469
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3931415922934417e-05,
+      "loss": 1.9047,
+      "step": 19470
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3930864630485053e-05,
+      "loss": 2.2212,
+      "step": 19471
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3930313323905162e-05,
+      "loss": 1.963,
+      "step": 19472
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3929762003196722e-05,
+      "loss": 1.6761,
+      "step": 19473
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.392921066836172e-05,
+      "loss": 1.8988,
+      "step": 19474
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3928659319402132e-05,
+      "loss": 1.9615,
+      "step": 19475
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3928107956319946e-05,
+      "loss": 1.7869,
+      "step": 19476
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3927556579117136e-05,
+      "loss": 1.7341,
+      "step": 19477
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3927005187795693e-05,
+      "loss": 1.9101,
+      "step": 19478
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3926453782357593e-05,
+      "loss": 2.028,
+      "step": 19479
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.392590236280482e-05,
+      "loss": 1.692,
+      "step": 19480
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3925350929139358e-05,
+      "loss": 1.6806,
+      "step": 19481
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3924799481363188e-05,
+      "loss": 1.9373,
+      "step": 19482
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3924248019478291e-05,
+      "loss": 2.0086,
+      "step": 19483
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3923696543486648e-05,
+      "loss": 1.9581,
+      "step": 19484
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3923145053390247e-05,
+      "loss": 1.9858,
+      "step": 19485
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3922593549191067e-05,
+      "loss": 2.0767,
+      "step": 19486
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.392204203089109e-05,
+      "loss": 1.7561,
+      "step": 19487
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3921490498492303e-05,
+      "loss": 1.8342,
+      "step": 19488
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3920938951996679e-05,
+      "loss": 1.8152,
+      "step": 19489
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3920387391406211e-05,
+      "loss": 1.9185,
+      "step": 19490
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3919835816722875e-05,
+      "loss": 1.9656,
+      "step": 19491
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.391928422794866e-05,
+      "loss": 1.7645,
+      "step": 19492
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3918732625085543e-05,
+      "loss": 1.9615,
+      "step": 19493
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3918181008135507e-05,
+      "loss": 1.7241,
+      "step": 19494
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3917629377100542e-05,
+      "loss": 1.5693,
+      "step": 19495
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3917077731982618e-05,
+      "loss": 1.9348,
+      "step": 19496
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3916526072783731e-05,
+      "loss": 1.8695,
+      "step": 19497
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3915974399505858e-05,
+      "loss": 1.7725,
+      "step": 19498
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3915422712150983e-05,
+      "loss": 1.8608,
+      "step": 19499
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3914871010721091e-05,
+      "loss": 1.9114,
+      "step": 19500
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3914319295218159e-05,
+      "loss": 2.0018,
+      "step": 19501
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3913767565644183e-05,
+      "loss": 1.8725,
+      "step": 19502
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.3913215822001132e-05,
+      "loss": 1.7874,
+      "step": 19503
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3912664064290996e-05,
+      "loss": 1.7738,
+      "step": 19504
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3912112292515762e-05,
+      "loss": 1.7491,
+      "step": 19505
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3911560506677404e-05,
+      "loss": 1.8357,
+      "step": 19506
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3911008706777913e-05,
+      "loss": 2.0668,
+      "step": 19507
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.3910456892819269e-05,
+      "loss": 1.8027,
+      "step": 19508
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3909905064803461e-05,
+      "loss": 1.797,
+      "step": 19509
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.3909353222732468e-05,
+      "loss": 1.985,
+      "step": 19510
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3908801366608273e-05,
+      "loss": 1.8224,
+      "step": 19511
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3908249496432867e-05,
+      "loss": 1.7473,
+      "step": 19512
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.390769761220822e-05,
+      "loss": 2.1081,
+      "step": 19513
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.390714571393633e-05,
+      "loss": 1.9975,
+      "step": 19514
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3906593801619173e-05,
+      "loss": 1.7418,
+      "step": 19515
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3906041875258734e-05,
+      "loss": 1.948,
+      "step": 19516
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.3905489934857e-05,
+      "loss": 1.1827,
+      "step": 19517
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3904937980415951e-05,
+      "loss": 1.7635,
+      "step": 19518
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3904386011937578e-05,
+      "loss": 1.7326,
+      "step": 19519
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3903834029423856e-05,
+      "loss": 1.8464,
+      "step": 19520
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3903282032876778e-05,
+      "loss": 1.9424,
+      "step": 19521
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.390273002229832e-05,
+      "loss": 1.6649,
+      "step": 19522
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3902177997690473e-05,
+      "loss": 2.1531,
+      "step": 19523
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.390162595905522e-05,
+      "loss": 1.9614,
+      "step": 19524
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3901073906394542e-05,
+      "loss": 1.8976,
+      "step": 19525
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3900521839710428e-05,
+      "loss": 1.8332,
+      "step": 19526
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.389996975900486e-05,
+      "loss": 1.6133,
+      "step": 19527
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.389941766427982e-05,
+      "loss": 1.8391,
+      "step": 19528
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.38988655555373e-05,
+      "loss": 1.9006,
+      "step": 19529
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3898313432779278e-05,
+      "loss": 1.5009,
+      "step": 19530
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3897761296007741e-05,
+      "loss": 1.8812,
+      "step": 19531
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3897209145224676e-05,
+      "loss": 2.1642,
+      "step": 19532
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3896656980432064e-05,
+      "loss": 1.7104,
+      "step": 19533
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.389610480163189e-05,
+      "loss": 2.4726,
+      "step": 19534
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3895552608826144e-05,
+      "loss": 1.9999,
+      "step": 19535
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3895000402016805e-05,
+      "loss": 2.1477,
+      "step": 19536
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3894448181205862e-05,
+      "loss": 1.7634,
+      "step": 19537
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.38938959463953e-05,
+      "loss": 1.7281,
+      "step": 19538
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.38933436975871e-05,
+      "loss": 2.0317,
+      "step": 19539
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3892791434783252e-05,
+      "loss": 1.7176,
+      "step": 19540
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3892239157985738e-05,
+      "loss": 2.024,
+      "step": 19541
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3891686867196547e-05,
+      "loss": 1.9308,
+      "step": 19542
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.389113456241766e-05,
+      "loss": 1.8411,
+      "step": 19543
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3890582243651065e-05,
+      "loss": 1.8458,
+      "step": 19544
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3890029910898747e-05,
+      "loss": 1.66,
+      "step": 19545
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3889477564162695e-05,
+      "loss": 1.7843,
+      "step": 19546
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3888925203444888e-05,
+      "loss": 1.8372,
+      "step": 19547
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3888372828747313e-05,
+      "loss": 1.745,
+      "step": 19548
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.388782044007196e-05,
+      "loss": 2.0434,
+      "step": 19549
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3887268037420809e-05,
+      "loss": 1.8941,
+      "step": 19550
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.388671562079585e-05,
+      "loss": 2.1325,
+      "step": 19551
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.388616319019907e-05,
+      "loss": 1.7866,
+      "step": 19552
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.388561074563245e-05,
+      "loss": 1.8388,
+      "step": 19553
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3885058287097977e-05,
+      "loss": 1.8353,
+      "step": 19554
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.388450581459764e-05,
+      "loss": 1.6169,
+      "step": 19555
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.3883953328133424e-05,
+      "loss": 1.7625,
+      "step": 19556
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3883400827707311e-05,
+      "loss": 1.7247,
+      "step": 19557
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3882848313321295e-05,
+      "loss": 1.5639,
+      "step": 19558
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3882295784977356e-05,
+      "loss": 1.818,
+      "step": 19559
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3881743242677481e-05,
+      "loss": 1.7489,
+      "step": 19560
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3881190686423658e-05,
+      "loss": 2.1163,
+      "step": 19561
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.388063811621787e-05,
+      "loss": 2.0408,
+      "step": 19562
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.388008553206211e-05,
+      "loss": 1.8444,
+      "step": 19563
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3879532933958352e-05,
+      "loss": 1.7029,
+      "step": 19564
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3878980321908598e-05,
+      "loss": 1.8922,
+      "step": 19565
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3878427695914827e-05,
+      "loss": 1.6986,
+      "step": 19566
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3877875055979021e-05,
+      "loss": 1.6096,
+      "step": 19567
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3877322402103178e-05,
+      "loss": 1.7936,
+      "step": 19568
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3876769734289273e-05,
+      "loss": 1.8545,
+      "step": 19569
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3876217052539298e-05,
+      "loss": 1.78,
+      "step": 19570
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3875664356855238e-05,
+      "loss": 1.714,
+      "step": 19571
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3875111647239085e-05,
+      "loss": 1.8301,
+      "step": 19572
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3874558923692818e-05,
+      "loss": 1.8299,
+      "step": 19573
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3874006186218431e-05,
+      "loss": 1.9433,
+      "step": 19574
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.3873453434817909e-05,
+      "loss": 2.2239,
+      "step": 19575
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3872900669493236e-05,
+      "loss": 1.7643,
+      "step": 19576
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3872347890246401e-05,
+      "loss": 2.1556,
+      "step": 19577
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3871795097079391e-05,
+      "loss": 1.828,
+      "step": 19578
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3871242289994196e-05,
+      "loss": 1.8158,
+      "step": 19579
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3870689468992796e-05,
+      "loss": 2.0514,
+      "step": 19580
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3870136634077182e-05,
+      "loss": 1.8842,
+      "step": 19581
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3869583785249345e-05,
+      "loss": 1.6802,
+      "step": 19582
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.3869030922511271e-05,
+      "loss": 2.3102,
+      "step": 19583
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3868478045864944e-05,
+      "loss": 1.7898,
+      "step": 19584
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3867925155312352e-05,
+      "loss": 1.6119,
+      "step": 19585
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3867372250855486e-05,
+      "loss": 1.835,
+      "step": 19586
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.386681933249633e-05,
+      "loss": 1.9469,
+      "step": 19587
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3866266400236874e-05,
+      "loss": 1.7009,
+      "step": 19588
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3865713454079107e-05,
+      "loss": 1.7778,
+      "step": 19589
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.386516049402501e-05,
+      "loss": 2.0236,
+      "step": 19590
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3864607520076576e-05,
+      "loss": 2.0214,
+      "step": 19591
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3864054532235791e-05,
+      "loss": 2.0116,
+      "step": 19592
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3863501530504648e-05,
+      "loss": 1.8139,
+      "step": 19593
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3862948514885127e-05,
+      "loss": 1.6705,
+      "step": 19594
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3862395485379221e-05,
+      "loss": 1.7584,
+      "step": 19595
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3861842441988917e-05,
+      "loss": 1.8605,
+      "step": 19596
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.0,
+      "learning_rate": 1.38612893847162e-05,
+      "loss": 2.157,
+      "step": 19597
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3860736313563063e-05,
+      "loss": 1.796,
+      "step": 19598
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.3860183228531492e-05,
+      "loss": 2.0749,
+      "step": 19599
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3859630129623475e-05,
+      "loss": 1.6214,
+      "step": 19600
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3859077016841e-05,
+      "loss": 2.0449,
+      "step": 19601
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3858523890186058e-05,
+      "loss": 1.8747,
+      "step": 19602
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3857970749660635e-05,
+      "loss": 1.8662,
+      "step": 19603
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3857417595266715e-05,
+      "loss": 1.9662,
+      "step": 19604
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3856864427006296e-05,
+      "loss": 1.9928,
+      "step": 19605
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.385631124488136e-05,
+      "loss": 1.708,
+      "step": 19606
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3855758048893895e-05,
+      "loss": 2.0696,
+      "step": 19607
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3855204839045894e-05,
+      "loss": 2.1277,
+      "step": 19608
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3854651615339343e-05,
+      "loss": 1.7663,
+      "step": 19609
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.385409837777623e-05,
+      "loss": 1.8358,
+      "step": 19610
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3853545126358542e-05,
+      "loss": 1.5949,
+      "step": 19611
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3852991861088276e-05,
+      "loss": 1.8002,
+      "step": 19612
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3852438581967414e-05,
+      "loss": 1.9338,
+      "step": 19613
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3851885288997946e-05,
+      "loss": 2.0279,
+      "step": 19614
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.3851331982181862e-05,
+      "loss": 1.9131,
+      "step": 19615
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3850778661521149e-05,
+      "loss": 1.7773,
+      "step": 19616
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.38502253270178e-05,
+      "loss": 1.9671,
+      "step": 19617
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3849671978673795e-05,
+      "loss": 1.6246,
+      "step": 19618
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.3849118616491137e-05,
+      "loss": 1.7939,
+      "step": 19619
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3848565240471806e-05,
+      "loss": 1.6824,
+      "step": 19620
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3848011850617791e-05,
+      "loss": 1.7057,
+      "step": 19621
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3847458446931084e-05,
+      "loss": 1.6299,
+      "step": 19622
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3846905029413675e-05,
+      "loss": 1.9209,
+      "step": 19623
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3846351598067551e-05,
+      "loss": 2.1969,
+      "step": 19624
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3845798152894702e-05,
+      "loss": 1.993,
+      "step": 19625
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3845244693897122e-05,
+      "loss": 1.6715,
+      "step": 19626
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3844691221076792e-05,
+      "loss": 1.8303,
+      "step": 19627
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3844137734435707e-05,
+      "loss": 1.9369,
+      "step": 19628
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.3843584233975857e-05,
+      "loss": 1.958,
+      "step": 19629
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3843030719699233e-05,
+      "loss": 2.1729,
+      "step": 19630
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.384247719160782e-05,
+      "loss": 1.7765,
+      "step": 19631
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3841923649703608e-05,
+      "loss": 1.7998,
+      "step": 19632
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3841370093988596e-05,
+      "loss": 1.7593,
+      "step": 19633
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.384081652446476e-05,
+      "loss": 1.8595,
+      "step": 19634
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.38402629411341e-05,
+      "loss": 2.0007,
+      "step": 19635
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3839709343998601e-05,
+      "loss": 1.6394,
+      "step": 19636
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3839155733060256e-05,
+      "loss": 1.6875,
+      "step": 19637
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3838602108321056e-05,
+      "loss": 1.6766,
+      "step": 19638
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3838048469782985e-05,
+      "loss": 1.9316,
+      "step": 19639
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3837494817448043e-05,
+      "loss": 1.8121,
+      "step": 19640
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3836941151318211e-05,
+      "loss": 1.7788,
+      "step": 19641
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3836387471395483e-05,
+      "loss": 2.2757,
+      "step": 19642
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3835833777681851e-05,
+      "loss": 2.1103,
+      "step": 19643
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3835280070179302e-05,
+      "loss": 2.144,
+      "step": 19644
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3834726348889831e-05,
+      "loss": 1.9491,
+      "step": 19645
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.3834172613815423e-05,
+      "loss": 2.0325,
+      "step": 19646
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3833618864958072e-05,
+      "loss": 1.5989,
+      "step": 19647
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3833065102319765e-05,
+      "loss": 1.7842,
+      "step": 19648
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3832511325902498e-05,
+      "loss": 1.5972,
+      "step": 19649
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.383195753570826e-05,
+      "loss": 1.8436,
+      "step": 19650
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.383140373173904e-05,
+      "loss": 1.6787,
+      "step": 19651
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3830849913996831e-05,
+      "loss": 2.0712,
+      "step": 19652
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.383029608248362e-05,
+      "loss": 1.6034,
+      "step": 19653
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3829742237201403e-05,
+      "loss": 2.0479,
+      "step": 19654
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3829188378152166e-05,
+      "loss": 1.6501,
+      "step": 19655
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3828634505337906e-05,
+      "loss": 2.033,
+      "step": 19656
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3828080618760606e-05,
+      "loss": 1.9022,
+      "step": 19657
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3827526718422264e-05,
+      "loss": 1.7846,
+      "step": 19658
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3826972804324867e-05,
+      "loss": 1.9385,
+      "step": 19659
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.382641887647041e-05,
+      "loss": 1.8326,
+      "step": 19660
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.382586493486088e-05,
+      "loss": 1.947,
+      "step": 19661
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.382531097949827e-05,
+      "loss": 1.4321,
+      "step": 19662
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3824757010384576e-05,
+      "loss": 1.9965,
+      "step": 19663
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3824203027521778e-05,
+      "loss": 2.3029,
+      "step": 19664
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3823649030911881e-05,
+      "loss": 2.3279,
+      "step": 19665
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3823095020556869e-05,
+      "loss": 1.6967,
+      "step": 19666
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.382254099645873e-05,
+      "loss": 1.8553,
+      "step": 19667
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3821986958619464e-05,
+      "loss": 2.1102,
+      "step": 19668
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3821432907041056e-05,
+      "loss": 1.8746,
+      "step": 19669
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.3820878841725505e-05,
+      "loss": 2.2137,
+      "step": 19670
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3820324762674794e-05,
+      "loss": 1.9677,
+      "step": 19671
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3819770669890922e-05,
+      "loss": 2.195,
+      "step": 19672
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3819216563375877e-05,
+      "loss": 1.7748,
+      "step": 19673
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3818662443131649e-05,
+      "loss": 1.895,
+      "step": 19674
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3818108309160233e-05,
+      "loss": 1.976,
+      "step": 19675
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3817554161463621e-05,
+      "loss": 1.7794,
+      "step": 19676
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.381700000004381e-05,
+      "loss": 1.9192,
+      "step": 19677
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.381644582490278e-05,
+      "loss": 2.0156,
+      "step": 19678
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3815891636042534e-05,
+      "loss": 1.8875,
+      "step": 19679
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.381533743346506e-05,
+      "loss": 1.9752,
+      "step": 19680
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3814783217172345e-05,
+      "loss": 2.2361,
+      "step": 19681
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3814228987166392e-05,
+      "loss": 1.8321,
+      "step": 19682
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3813674743449187e-05,
+      "loss": 1.5569,
+      "step": 19683
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3813120486022719e-05,
+      "loss": 1.9854,
+      "step": 19684
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.381256621488899e-05,
+      "loss": 1.8055,
+      "step": 19685
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3812011930049982e-05,
+      "loss": 2.0896,
+      "step": 19686
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3811457631507698e-05,
+      "loss": 1.9696,
+      "step": 19687
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3810903319264121e-05,
+      "loss": 1.9459,
+      "step": 19688
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.381034899332125e-05,
+      "loss": 1.7233,
+      "step": 19689
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3809794653681075e-05,
+      "loss": 1.7848,
+      "step": 19690
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.380924030034559e-05,
+      "loss": 1.8184,
+      "step": 19691
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3808685933316786e-05,
+      "loss": 1.6292,
+      "step": 19692
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3808131552596657e-05,
+      "loss": 2.0134,
+      "step": 19693
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3807577158187196e-05,
+      "loss": 2.0366,
+      "step": 19694
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3807022750090393e-05,
+      "loss": 1.7571,
+      "step": 19695
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3806468328308244e-05,
+      "loss": 1.9549,
+      "step": 19696
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3805913892842746e-05,
+      "loss": 1.9898,
+      "step": 19697
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3805359443695885e-05,
+      "loss": 2.0538,
+      "step": 19698
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.125,
+      "learning_rate": 1.3804804980869657e-05,
+      "loss": 1.7173,
+      "step": 19699
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3804250504366054e-05,
+      "loss": 2.0178,
+      "step": 19700
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3803696014187072e-05,
+      "loss": 1.9475,
+      "step": 19701
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3803141510334698e-05,
+      "loss": 1.8883,
+      "step": 19702
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3802586992810935e-05,
+      "loss": 1.8833,
+      "step": 19703
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3802032461617767e-05,
+      "loss": 1.9452,
+      "step": 19704
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3801477916757193e-05,
+      "loss": 2.1068,
+      "step": 19705
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3800923358231206e-05,
+      "loss": 1.8089,
+      "step": 19706
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3800368786041798e-05,
+      "loss": 1.8898,
+      "step": 19707
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3799814200190965e-05,
+      "loss": 1.9066,
+      "step": 19708
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3799259600680692e-05,
+      "loss": 1.9537,
+      "step": 19709
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3798704987512985e-05,
+      "loss": 1.7139,
+      "step": 19710
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3798150360689829e-05,
+      "loss": 2.1837,
+      "step": 19711
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3797595720213224e-05,
+      "loss": 1.9637,
+      "step": 19712
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3797041066085157e-05,
+      "loss": 1.9938,
+      "step": 19713
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3796486398307627e-05,
+      "loss": 1.671,
+      "step": 19714
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3795931716882628e-05,
+      "loss": 2.0334,
+      "step": 19715
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3795377021812146e-05,
+      "loss": 2.0432,
+      "step": 19716
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3794822313098188e-05,
+      "loss": 2.1227,
+      "step": 19717
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3794267590742738e-05,
+      "loss": 2.039,
+      "step": 19718
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3793712854747794e-05,
+      "loss": 1.9403,
+      "step": 19719
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.379315810511535e-05,
+      "loss": 1.7196,
+      "step": 19720
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3792603341847399e-05,
+      "loss": 1.9503,
+      "step": 19721
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3792048564945937e-05,
+      "loss": 1.6682,
+      "step": 19722
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3791493774412954e-05,
+      "loss": 2.1426,
+      "step": 19723
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3790938970250452e-05,
+      "loss": 1.9956,
+      "step": 19724
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3790384152460417e-05,
+      "loss": 1.8076,
+      "step": 19725
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.378982932104485e-05,
+      "loss": 2.0461,
+      "step": 19726
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.378927447600574e-05,
+      "loss": 1.699,
+      "step": 19727
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3788719617345087e-05,
+      "loss": 2.1781,
+      "step": 19728
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3788164745064884e-05,
+      "loss": 1.6936,
+      "step": 19729
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3787609859167124e-05,
+      "loss": 1.9088,
+      "step": 19730
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.37870549596538e-05,
+      "loss": 1.8933,
+      "step": 19731
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3786500046526909e-05,
+      "loss": 1.8731,
+      "step": 19732
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3785945119788449e-05,
+      "loss": 1.9228,
+      "step": 19733
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3785390179440409e-05,
+      "loss": 1.9057,
+      "step": 19734
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3784835225484787e-05,
+      "loss": 2.0993,
+      "step": 19735
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3784280257923577e-05,
+      "loss": 1.9497,
+      "step": 19736
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3783725276758778e-05,
+      "loss": 1.7792,
+      "step": 19737
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3783170281992378e-05,
+      "loss": 1.6299,
+      "step": 19738
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3782615273626374e-05,
+      "loss": 1.7435,
+      "step": 19739
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3782060251662767e-05,
+      "loss": 2.1084,
+      "step": 19740
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3781505216103543e-05,
+      "loss": 1.8415,
+      "step": 19741
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3780950166950706e-05,
+      "loss": 1.7234,
+      "step": 19742
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3780395104206246e-05,
+      "loss": 1.7571,
+      "step": 19743
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.377984002787216e-05,
+      "loss": 1.8214,
+      "step": 19744
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3779284937950443e-05,
+      "loss": 1.9423,
+      "step": 19745
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3778729834443088e-05,
+      "loss": 1.4765,
+      "step": 19746
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3778174717352098e-05,
+      "loss": 1.9834,
+      "step": 19747
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3777619586679458e-05,
+      "loss": 1.9394,
+      "step": 19748
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3777064442427172e-05,
+      "loss": 1.8444,
+      "step": 19749
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3776509284597235e-05,
+      "loss": 2.0248,
+      "step": 19750
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3775954113191634e-05,
+      "loss": 2.1209,
+      "step": 19751
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3775398928212376e-05,
+      "loss": 1.9207,
+      "step": 19752
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3774843729661447e-05,
+      "loss": 1.8115,
+      "step": 19753
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3774288517540853e-05,
+      "loss": 1.7361,
+      "step": 19754
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3773733291852578e-05,
+      "loss": 1.945,
+      "step": 19755
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.377317805259863e-05,
+      "loss": 1.8519,
+      "step": 19756
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3772622799780999e-05,
+      "loss": 2.0135,
+      "step": 19757
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3772067533401678e-05,
+      "loss": 1.8471,
+      "step": 19758
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3771512253462666e-05,
+      "loss": 2.021,
+      "step": 19759
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3770956959965962e-05,
+      "loss": 1.8874,
+      "step": 19760
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3770401652913559e-05,
+      "loss": 1.7173,
+      "step": 19761
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3769846332307453e-05,
+      "loss": 1.9113,
+      "step": 19762
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.376929099814964e-05,
+      "loss": 1.8955,
+      "step": 19763
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3768735650442123e-05,
+      "loss": 1.6485,
+      "step": 19764
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3768180289186886e-05,
+      "loss": 2.0753,
+      "step": 19765
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3767624914385933e-05,
+      "loss": 2.0888,
+      "step": 19766
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3767069526041261e-05,
+      "loss": 2.0277,
+      "step": 19767
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3766514124154867e-05,
+      "loss": 2.3253,
+      "step": 19768
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3765958708728742e-05,
+      "loss": 1.87,
+      "step": 19769
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3765403279764886e-05,
+      "loss": 1.7645,
+      "step": 19770
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3764847837265298e-05,
+      "loss": 1.8353,
+      "step": 19771
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.376429238123197e-05,
+      "loss": 2.0001,
+      "step": 19772
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3763736911666903e-05,
+      "loss": 1.9907,
+      "step": 19773
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3763181428572092e-05,
+      "loss": 2.0763,
+      "step": 19774
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3762625931949534e-05,
+      "loss": 1.7296,
+      "step": 19775
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3762070421801223e-05,
+      "loss": 2.0831,
+      "step": 19776
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3761514898129163e-05,
+      "loss": 1.8944,
+      "step": 19777
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3760959360935342e-05,
+      "loss": 1.7658,
+      "step": 19778
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.376040381022176e-05,
+      "loss": 2.108,
+      "step": 19779
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3759848245990419e-05,
+      "loss": 1.9551,
+      "step": 19780
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3759292668243313e-05,
+      "loss": 2.0809,
+      "step": 19781
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.375873707698244e-05,
+      "loss": 1.9478,
+      "step": 19782
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3758181472209795e-05,
+      "loss": 2.0609,
+      "step": 19783
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3757625853927377e-05,
+      "loss": 2.0166,
+      "step": 19784
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.3757070222137181e-05,
+      "loss": 1.726,
+      "step": 19785
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3756514576841205e-05,
+      "loss": 1.9238,
+      "step": 19786
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3755958918041452e-05,
+      "loss": 1.6214,
+      "step": 19787
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3755403245739912e-05,
+      "loss": 1.8453,
+      "step": 19788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3754847559938585e-05,
+      "loss": 1.8687,
+      "step": 19789
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3754291860639471e-05,
+      "loss": 1.8182,
+      "step": 19790
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3753736147844563e-05,
+      "loss": 1.9629,
+      "step": 19791
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3753180421555863e-05,
+      "loss": 1.9026,
+      "step": 19792
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3752624681775364e-05,
+      "loss": 1.7486,
+      "step": 19793
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3752068928505073e-05,
+      "loss": 2.1306,
+      "step": 19794
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3751513161746975e-05,
+      "loss": 1.6985,
+      "step": 19795
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3750957381503079e-05,
+      "loss": 1.7957,
+      "step": 19796
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3750401587775375e-05,
+      "loss": 1.9047,
+      "step": 19797
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3749845780565867e-05,
+      "loss": 2.1648,
+      "step": 19798
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3749289959876549e-05,
+      "loss": 1.8301,
+      "step": 19799
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3748734125709419e-05,
+      "loss": 1.9179,
+      "step": 19800
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.374817827806648e-05,
+      "loss": 1.7696,
+      "step": 19801
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3747622416949722e-05,
+      "loss": 1.6591,
+      "step": 19802
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3747066542361149e-05,
+      "loss": 1.8163,
+      "step": 19803
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.374651065430276e-05,
+      "loss": 1.5063,
+      "step": 19804
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3745954752776548e-05,
+      "loss": 1.9542,
+      "step": 19805
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3745398837784518e-05,
+      "loss": 2.2479,
+      "step": 19806
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3744842909328662e-05,
+      "loss": 1.6462,
+      "step": 19807
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3744286967410988e-05,
+      "loss": 2.1914,
+      "step": 19808
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3743731012033479e-05,
+      "loss": 2.0224,
+      "step": 19809
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3743175043198147e-05,
+      "loss": 1.4968,
+      "step": 19810
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3742619060906985e-05,
+      "loss": 1.6922,
+      "step": 19811
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3742063065161993e-05,
+      "loss": 2.0608,
+      "step": 19812
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3741507055965169e-05,
+      "loss": 1.9905,
+      "step": 19813
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3740951033318512e-05,
+      "loss": 1.6494,
+      "step": 19814
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.374039499722402e-05,
+      "loss": 1.6726,
+      "step": 19815
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3739838947683693e-05,
+      "loss": 1.8848,
+      "step": 19816
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.373928288469953e-05,
+      "loss": 2.045,
+      "step": 19817
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3738726808273528e-05,
+      "loss": 2.0116,
+      "step": 19818
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3738170718407689e-05,
+      "loss": 1.8939,
+      "step": 19819
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3737614615104012e-05,
+      "loss": 1.7432,
+      "step": 19820
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.373705849836449e-05,
+      "loss": 1.7489,
+      "step": 19821
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.373650236819113e-05,
+      "loss": 1.7674,
+      "step": 19822
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3735946224585923e-05,
+      "loss": 1.8237,
+      "step": 19823
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3735390067550878e-05,
+      "loss": 2.0308,
+      "step": 19824
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3734833897087986e-05,
+      "loss": 1.5299,
+      "step": 19825
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3734277713199252e-05,
+      "loss": 2.1361,
+      "step": 19826
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3733721515886671e-05,
+      "loss": 1.6787,
+      "step": 19827
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3733165305152242e-05,
+      "loss": 1.9171,
+      "step": 19828
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3732609080997972e-05,
+      "loss": 1.6869,
+      "step": 19829
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.373205284342585e-05,
+      "loss": 1.556,
+      "step": 19830
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3731496592437886e-05,
+      "loss": 1.6768,
+      "step": 19831
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3730940328036066e-05,
+      "loss": 1.7973,
+      "step": 19832
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3730384050222406e-05,
+      "loss": 1.8124,
+      "step": 19833
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3729827758998894e-05,
+      "loss": 1.8428,
+      "step": 19834
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3729271454367533e-05,
+      "loss": 1.8327,
+      "step": 19835
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3728715136330324e-05,
+      "loss": 1.8204,
+      "step": 19836
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3728158804889264e-05,
+      "loss": 1.606,
+      "step": 19837
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.372760246004636e-05,
+      "loss": 1.9447,
+      "step": 19838
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.37270461018036e-05,
+      "loss": 1.9382,
+      "step": 19839
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3726489730162995e-05,
+      "loss": 1.7926,
+      "step": 19840
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.3725933345126541e-05,
+      "loss": 1.6043,
+      "step": 19841
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3725376946696233e-05,
+      "loss": 2.2198,
+      "step": 19842
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3724820534874082e-05,
+      "loss": 1.9463,
+      "step": 19843
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3724264109662076e-05,
+      "loss": 1.7369,
+      "step": 19844
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3723707671062228e-05,
+      "loss": 2.0023,
+      "step": 19845
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3723151219076526e-05,
+      "loss": 2.1875,
+      "step": 19846
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3722594753706977e-05,
+      "loss": 1.6581,
+      "step": 19847
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3722038274955584e-05,
+      "loss": 1.7424,
+      "step": 19848
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.372148178282434e-05,
+      "loss": 1.878,
+      "step": 19849
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3720925277315249e-05,
+      "loss": 1.7188,
+      "step": 19850
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3720368758430313e-05,
+      "loss": 2.0351,
+      "step": 19851
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3719812226171532e-05,
+      "loss": 2.0454,
+      "step": 19852
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3719255680540904e-05,
+      "loss": 1.8176,
+      "step": 19853
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.371869912154043e-05,
+      "loss": 1.8177,
+      "step": 19854
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3718142549172115e-05,
+      "loss": 1.8815,
+      "step": 19855
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.3717585963437956e-05,
+      "loss": 2.2542,
+      "step": 19856
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3717029364339954e-05,
+      "loss": 1.8798,
+      "step": 19857
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.371647275188011e-05,
+      "loss": 1.6584,
+      "step": 19858
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3715916126060426e-05,
+      "loss": 1.7524,
+      "step": 19859
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3715359486882904e-05,
+      "loss": 2.0628,
+      "step": 19860
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3714802834349544e-05,
+      "loss": 1.9145,
+      "step": 19861
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3714246168462343e-05,
+      "loss": 2.0681,
+      "step": 19862
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3713689489223305e-05,
+      "loss": 1.8388,
+      "step": 19863
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3713132796634434e-05,
+      "loss": 1.9869,
+      "step": 19864
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3712576090697727e-05,
+      "loss": 1.7235,
+      "step": 19865
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3712019371415188e-05,
+      "loss": 1.8614,
+      "step": 19866
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3711462638788817e-05,
+      "loss": 1.709,
+      "step": 19867
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3710905892820616e-05,
+      "loss": 1.6861,
+      "step": 19868
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3710349133512586e-05,
+      "loss": 1.7856,
+      "step": 19869
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3709792360866727e-05,
+      "loss": 1.9581,
+      "step": 19870
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3709235574885044e-05,
+      "loss": 1.9813,
+      "step": 19871
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3708678775569534e-05,
+      "loss": 1.5444,
+      "step": 19872
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3708121962922204e-05,
+      "loss": 1.9004,
+      "step": 19873
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.370756513694505e-05,
+      "loss": 2.126,
+      "step": 19874
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3707008297640076e-05,
+      "loss": 2.1036,
+      "step": 19875
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3706451445009283e-05,
+      "loss": 2.0431,
+      "step": 19876
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3705894579054675e-05,
+      "loss": 2.0422,
+      "step": 19877
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3705337699778253e-05,
+      "loss": 2.0066,
+      "step": 19878
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3704780807182019e-05,
+      "loss": 1.8456,
+      "step": 19879
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3704223901267972e-05,
+      "loss": 1.6421,
+      "step": 19880
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3703666982038118e-05,
+      "loss": 1.6419,
+      "step": 19881
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3703110049494454e-05,
+      "loss": 1.6969,
+      "step": 19882
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3702553103638989e-05,
+      "loss": 1.753,
+      "step": 19883
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3701996144473717e-05,
+      "loss": 1.8535,
+      "step": 19884
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.370143917200065e-05,
+      "loss": 1.9212,
+      "step": 19885
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3700882186221778e-05,
+      "loss": 1.5114,
+      "step": 19886
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3700325187139114e-05,
+      "loss": 1.8932,
+      "step": 19887
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3699768174754655e-05,
+      "loss": 1.947,
+      "step": 19888
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3699211149070407e-05,
+      "loss": 1.8587,
+      "step": 19889
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3698654110088365e-05,
+      "loss": 1.9092,
+      "step": 19890
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3698097057810538e-05,
+      "loss": 1.9344,
+      "step": 19891
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3697539992238928e-05,
+      "loss": 1.8833,
+      "step": 19892
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3696982913375534e-05,
+      "loss": 2.4146,
+      "step": 19893
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3696425821222362e-05,
+      "loss": 1.798,
+      "step": 19894
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3695868715781416e-05,
+      "loss": 1.9021,
+      "step": 19895
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3695311597054693e-05,
+      "loss": 2.1837,
+      "step": 19896
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3694754465044197e-05,
+      "loss": 1.9236,
+      "step": 19897
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3694197319751936e-05,
+      "loss": 2.0744,
+      "step": 19898
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3693640161179909e-05,
+      "loss": 2.0031,
+      "step": 19899
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3693082989330115e-05,
+      "loss": 1.7929,
+      "step": 19900
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3692525804204566e-05,
+      "loss": 1.9187,
+      "step": 19901
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3691968605805256e-05,
+      "loss": 1.7797,
+      "step": 19902
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3691411394134196e-05,
+      "loss": 1.9445,
+      "step": 19903
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3690854169193383e-05,
+      "loss": 1.9567,
+      "step": 19904
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.3690296930984821e-05,
+      "loss": 1.5952,
+      "step": 19905
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3689739679510517e-05,
+      "loss": 1.7889,
+      "step": 19906
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3689182414772468e-05,
+      "loss": 1.6791,
+      "step": 19907
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3688625136772685e-05,
+      "loss": 1.9019,
+      "step": 19908
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3688067845513162e-05,
+      "loss": 1.9602,
+      "step": 19909
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3687510540995911e-05,
+      "loss": 1.9012,
+      "step": 19910
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3686953223222932e-05,
+      "loss": 2.0773,
+      "step": 19911
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3686395892196225e-05,
+      "loss": 1.8088,
+      "step": 19912
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3685838547917798e-05,
+      "loss": 1.985,
+      "step": 19913
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3685281190389653e-05,
+      "loss": 1.8526,
+      "step": 19914
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3684723819613797e-05,
+      "loss": 1.8922,
+      "step": 19915
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3684166435592223e-05,
+      "loss": 1.9074,
+      "step": 19916
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3683609038326948e-05,
+      "loss": 2.1884,
+      "step": 19917
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.368305162781997e-05,
+      "loss": 1.9737,
+      "step": 19918
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3682494204073288e-05,
+      "loss": 2.038,
+      "step": 19919
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3681936767088913e-05,
+      "loss": 1.9267,
+      "step": 19920
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3681379316868844e-05,
+      "loss": 1.7477,
+      "step": 19921
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3680821853415092e-05,
+      "loss": 2.0458,
+      "step": 19922
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.368026437672965e-05,
+      "loss": 2.089,
+      "step": 19923
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.367970688681453e-05,
+      "loss": 2.1345,
+      "step": 19924
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3679149383671737e-05,
+      "loss": 1.5421,
+      "step": 19925
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3678591867303267e-05,
+      "loss": 1.496,
+      "step": 19926
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3678034337711133e-05,
+      "loss": 1.7126,
+      "step": 19927
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.367747679489733e-05,
+      "loss": 2.0269,
+      "step": 19928
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3676919238863875e-05,
+      "loss": 1.9307,
+      "step": 19929
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3676361669612764e-05,
+      "loss": 1.9318,
+      "step": 19930
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3675804087145997e-05,
+      "loss": 2.2661,
+      "step": 19931
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.3675246491465589e-05,
+      "loss": 1.8092,
+      "step": 19932
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3674688882573532e-05,
+      "loss": 1.652,
+      "step": 19933
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3674131260471843e-05,
+      "loss": 1.7455,
+      "step": 19934
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3673573625162521e-05,
+      "loss": 2.0209,
+      "step": 19935
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.367301597664757e-05,
+      "loss": 1.7213,
+      "step": 19936
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3672458314928996e-05,
+      "loss": 2.1291,
+      "step": 19937
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.36719006400088e-05,
+      "loss": 1.903,
+      "step": 19938
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3671342951888993e-05,
+      "loss": 1.6766,
+      "step": 19939
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3670785250571572e-05,
+      "loss": 1.8506,
+      "step": 19940
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3670227536058549e-05,
+      "loss": 2.0415,
+      "step": 19941
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3669669808351926e-05,
+      "loss": 1.9014,
+      "step": 19942
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3669112067453706e-05,
+      "loss": 1.9758,
+      "step": 19943
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3668554313365898e-05,
+      "loss": 1.6906,
+      "step": 19944
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3667996546090506e-05,
+      "loss": 2.1405,
+      "step": 19945
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3667438765629533e-05,
+      "loss": 1.8091,
+      "step": 19946
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.366688097198498e-05,
+      "loss": 1.5622,
+      "step": 19947
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3666323165158863e-05,
+      "loss": 1.8713,
+      "step": 19948
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.366576534515318e-05,
+      "loss": 2.0739,
+      "step": 19949
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3665207511969936e-05,
+      "loss": 1.8574,
+      "step": 19950
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.366464966561114e-05,
+      "loss": 2.04,
+      "step": 19951
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3664091806078795e-05,
+      "loss": 1.8106,
+      "step": 19952
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3663533933374904e-05,
+      "loss": 1.9088,
+      "step": 19953
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3662976047501474e-05,
+      "loss": 1.5786,
+      "step": 19954
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3662418148460517e-05,
+      "loss": 1.8963,
+      "step": 19955
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.3661860236254027e-05,
+      "loss": 1.8778,
+      "step": 19956
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3661302310884019e-05,
+      "loss": 1.9279,
+      "step": 19957
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3660744372352495e-05,
+      "loss": 1.9191,
+      "step": 19958
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.366018642066146e-05,
+      "loss": 2.1614,
+      "step": 19959
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3659628455812919e-05,
+      "loss": 1.8654,
+      "step": 19960
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.365907047780888e-05,
+      "loss": 1.9167,
+      "step": 19961
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.365851248665135e-05,
+      "loss": 1.9223,
+      "step": 19962
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3657954482342329e-05,
+      "loss": 1.9075,
+      "step": 19963
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.365739646488383e-05,
+      "loss": 2.3699,
+      "step": 19964
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3656838434277854e-05,
+      "loss": 2.0505,
+      "step": 19965
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3656280390526411e-05,
+      "loss": 1.6,
+      "step": 19966
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3655722333631503e-05,
+      "loss": 1.7448,
+      "step": 19967
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3655164263595136e-05,
+      "loss": 1.8052,
+      "step": 19968
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3654606180419322e-05,
+      "loss": 1.8111,
+      "step": 19969
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3654048084106058e-05,
+      "loss": 1.7721,
+      "step": 19970
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3653489974657359e-05,
+      "loss": 1.7585,
+      "step": 19971
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3652931852075227e-05,
+      "loss": 2.0514,
+      "step": 19972
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3652373716361667e-05,
+      "loss": 1.9892,
+      "step": 19973
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.365181556751869e-05,
+      "loss": 1.7602,
+      "step": 19974
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3651257405548294e-05,
+      "loss": 1.7657,
+      "step": 19975
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.36506992304525e-05,
+      "loss": 1.5195,
+      "step": 19976
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3650141042233299e-05,
+      "loss": 1.6746,
+      "step": 19977
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3649582840892705e-05,
+      "loss": 2.0097,
+      "step": 19978
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3649024626432725e-05,
+      "loss": 1.9452,
+      "step": 19979
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3648466398855367e-05,
+      "loss": 2.2265,
+      "step": 19980
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3647908158162632e-05,
+      "loss": 1.8448,
+      "step": 19981
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.364734990435653e-05,
+      "loss": 2.0668,
+      "step": 19982
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.364679163743907e-05,
+      "loss": 1.8875,
+      "step": 19983
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.364623335741225e-05,
+      "loss": 1.7483,
+      "step": 19984
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.364567506427809e-05,
+      "loss": 2.0078,
+      "step": 19985
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.364511675803859e-05,
+      "loss": 2.319,
+      "step": 19986
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3644558438695755e-05,
+      "loss": 1.7085,
+      "step": 19987
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3644000106251593e-05,
+      "loss": 2.1477,
+      "step": 19988
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3643441760708116e-05,
+      "loss": 2.1756,
+      "step": 19989
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3642883402067326e-05,
+      "loss": 2.0948,
+      "step": 19990
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3642325030331232e-05,
+      "loss": 1.9866,
+      "step": 19991
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.3641766645501841e-05,
+      "loss": 2.2072,
+      "step": 19992
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3641208247581159e-05,
+      "loss": 1.887,
+      "step": 19993
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3640649836571194e-05,
+      "loss": 1.9842,
+      "step": 19994
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3640091412473955e-05,
+      "loss": 2.5105,
+      "step": 19995
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.363953297529145e-05,
+      "loss": 1.7662,
+      "step": 19996
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3638974525025681e-05,
+      "loss": 1.7111,
+      "step": 19997
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.3638416061678661e-05,
+      "loss": 2.1825,
+      "step": 19998
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.3637857585252396e-05,
+      "loss": 1.8281,
+      "step": 19999
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3637299095748892e-05,
+      "loss": 1.6051,
+      "step": 20000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3636740593170158e-05,
+      "loss": 1.9043,
+      "step": 20001
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3636182077518208e-05,
+      "loss": 1.9113,
+      "step": 20002
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3635623548795037e-05,
+      "loss": 1.9426,
+      "step": 20003
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3635065007002659e-05,
+      "loss": 1.9147,
+      "step": 20004
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3634506452143081e-05,
+      "loss": 1.744,
+      "step": 20005
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3633947884218317e-05,
+      "loss": 1.1595,
+      "step": 20006
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3633389303230364e-05,
+      "loss": 1.738,
+      "step": 20007
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.375,
+      "learning_rate": 1.363283070918124e-05,
+      "loss": 1.8629,
+      "step": 20008
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.363227210207295e-05,
+      "loss": 1.9115,
+      "step": 20009
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3631713481907495e-05,
+      "loss": 1.7186,
+      "step": 20010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3631154848686891e-05,
+      "loss": 1.7259,
+      "step": 20011
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3630596202413145e-05,
+      "loss": 1.9694,
+      "step": 20012
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3630037543088265e-05,
+      "loss": 1.5849,
+      "step": 20013
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.362947887071426e-05,
+      "loss": 1.7089,
+      "step": 20014
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.362892018529313e-05,
+      "loss": 2.1659,
+      "step": 20015
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3628361486826897e-05,
+      "loss": 2.0361,
+      "step": 20016
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.362780277531756e-05,
+      "loss": 1.8297,
+      "step": 20017
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.362724405076713e-05,
+      "loss": 1.9238,
+      "step": 20018
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3626685313177617e-05,
+      "loss": 2.0016,
+      "step": 20019
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3626126562551026e-05,
+      "loss": 1.9873,
+      "step": 20020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3625567798889367e-05,
+      "loss": 1.9234,
+      "step": 20021
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.362500902219465e-05,
+      "loss": 1.9206,
+      "step": 20022
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3624450232468887e-05,
+      "loss": 1.5823,
+      "step": 20023
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3623891429714075e-05,
+      "loss": 1.9495,
+      "step": 20024
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3623332613932236e-05,
+      "loss": 1.7344,
+      "step": 20025
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.3622773785125371e-05,
+      "loss": 2.2021,
+      "step": 20026
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.362221494329549e-05,
+      "loss": 1.8436,
+      "step": 20027
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3621656088444605e-05,
+      "loss": 2.0677,
+      "step": 20028
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3621097220574724e-05,
+      "loss": 1.8325,
+      "step": 20029
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3620538339687852e-05,
+      "loss": 1.7806,
+      "step": 20030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3619979445786e-05,
+      "loss": 1.483,
+      "step": 20031
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.361942053887118e-05,
+      "loss": 1.9068,
+      "step": 20032
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.36188616189454e-05,
+      "loss": 2.0114,
+      "step": 20033
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3618302686010668e-05,
+      "loss": 1.8029,
+      "step": 20034
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.361774374006899e-05,
+      "loss": 2.0729,
+      "step": 20035
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3617184781122383e-05,
+      "loss": 1.7152,
+      "step": 20036
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3616625809172852e-05,
+      "loss": 2.0561,
+      "step": 20037
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3616066824222403e-05,
+      "loss": 1.5811,
+      "step": 20038
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3615507826273055e-05,
+      "loss": 1.8242,
+      "step": 20039
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3614948815326805e-05,
+      "loss": 1.756,
+      "step": 20040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3614389791385672e-05,
+      "loss": 2.4798,
+      "step": 20041
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3613830754451662e-05,
+      "loss": 1.5826,
+      "step": 20042
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3613271704526785e-05,
+      "loss": 1.77,
+      "step": 20043
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3612712641613053e-05,
+      "loss": 1.8599,
+      "step": 20044
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.361215356571247e-05,
+      "loss": 2.2259,
+      "step": 20045
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.361159447682705e-05,
+      "loss": 1.7001,
+      "step": 20046
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3611035374958804e-05,
+      "loss": 1.8278,
+      "step": 20047
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3610476260109738e-05,
+      "loss": 1.8416,
+      "step": 20048
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3609917132281862e-05,
+      "loss": 1.8201,
+      "step": 20049
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.360935799147719e-05,
+      "loss": 1.6272,
+      "step": 20050
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.360879883769773e-05,
+      "loss": 1.8266,
+      "step": 20051
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3608239670945488e-05,
+      "loss": 1.8551,
+      "step": 20052
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3607680491222484e-05,
+      "loss": 2.1025,
+      "step": 20053
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3607121298530716e-05,
+      "loss": 2.0012,
+      "step": 20054
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3606562092872205e-05,
+      "loss": 1.5623,
+      "step": 20055
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3606002874248953e-05,
+      "loss": 2.0168,
+      "step": 20056
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3605443642662975e-05,
+      "loss": 1.9194,
+      "step": 20057
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3604884398116278e-05,
+      "loss": 2.0208,
+      "step": 20058
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3604325140610872e-05,
+      "loss": 1.6571,
+      "step": 20059
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3603765870148776e-05,
+      "loss": 1.7904,
+      "step": 20060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.360320658673199e-05,
+      "loss": 1.5279,
+      "step": 20061
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3602647290362528e-05,
+      "loss": 1.7408,
+      "step": 20062
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3602087981042402e-05,
+      "loss": 1.6086,
+      "step": 20063
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3601528658773622e-05,
+      "loss": 1.8915,
+      "step": 20064
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3600969323558198e-05,
+      "loss": 1.6441,
+      "step": 20065
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3600409975398142e-05,
+      "loss": 1.842,
+      "step": 20066
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.359985061429546e-05,
+      "loss": 1.8505,
+      "step": 20067
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3599291240252168e-05,
+      "loss": 1.7789,
+      "step": 20068
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3598731853270275e-05,
+      "loss": 1.9282,
+      "step": 20069
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3598172453351794e-05,
+      "loss": 2.0765,
+      "step": 20070
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3597613040498731e-05,
+      "loss": 1.8901,
+      "step": 20071
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3597053614713103e-05,
+      "loss": 1.7547,
+      "step": 20072
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3596494175996917e-05,
+      "loss": 1.985,
+      "step": 20073
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3595934724352184e-05,
+      "loss": 1.5932,
+      "step": 20074
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3595375259780916e-05,
+      "loss": 1.9462,
+      "step": 20075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3594815782285127e-05,
+      "loss": 1.6527,
+      "step": 20076
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3594256291866823e-05,
+      "loss": 1.7375,
+      "step": 20077
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3593696788528016e-05,
+      "loss": 1.7215,
+      "step": 20078
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3593137272270723e-05,
+      "loss": 2.0506,
+      "step": 20079
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.359257774309695e-05,
+      "loss": 1.8409,
+      "step": 20080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3592018201008708e-05,
+      "loss": 1.8939,
+      "step": 20081
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3591458646008009e-05,
+      "loss": 1.7903,
+      "step": 20082
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3590899078096871e-05,
+      "loss": 1.807,
+      "step": 20083
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3590339497277295e-05,
+      "loss": 2.0735,
+      "step": 20084
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.35897799035513e-05,
+      "loss": 1.8779,
+      "step": 20085
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3589220296920896e-05,
+      "loss": 1.7488,
+      "step": 20086
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3588660677388087e-05,
+      "loss": 1.7921,
+      "step": 20087
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3588101044954897e-05,
+      "loss": 1.6133,
+      "step": 20088
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3587541399623331e-05,
+      "loss": 1.9128,
+      "step": 20089
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3586981741395406e-05,
+      "loss": 2.1044,
+      "step": 20090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3586422070273129e-05,
+      "loss": 1.6701,
+      "step": 20091
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3585862386258509e-05,
+      "loss": 2.0312,
+      "step": 20092
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3585302689353564e-05,
+      "loss": 1.6472,
+      "step": 20093
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3584742979560302e-05,
+      "loss": 1.6261,
+      "step": 20094
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.3584183256880739e-05,
+      "loss": 2.2535,
+      "step": 20095
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.358362352131688e-05,
+      "loss": 1.6297,
+      "step": 20096
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3583063772870748e-05,
+      "loss": 2.0778,
+      "step": 20097
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.358250401154435e-05,
+      "loss": 1.7948,
+      "step": 20098
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.358194423733969e-05,
+      "loss": 1.8238,
+      "step": 20099
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3581384450258796e-05,
+      "loss": 1.6647,
+      "step": 20100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3580824650303666e-05,
+      "loss": 2.0088,
+      "step": 20101
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3580264837476321e-05,
+      "loss": 1.9395,
+      "step": 20102
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3579705011778767e-05,
+      "loss": 2.0194,
+      "step": 20103
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3579145173213025e-05,
+      "loss": 1.8142,
+      "step": 20104
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.35785853217811e-05,
+      "loss": 1.801,
+      "step": 20105
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3578025457485005e-05,
+      "loss": 2.018,
+      "step": 20106
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.375,
+      "learning_rate": 1.357746558032676e-05,
+      "loss": 1.8937,
+      "step": 20107
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.3576905690308366e-05,
+      "loss": 2.2319,
+      "step": 20108
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3576345787431845e-05,
+      "loss": 1.7689,
+      "step": 20109
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3575785871699208e-05,
+      "loss": 1.848,
+      "step": 20110
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3575225943112464e-05,
+      "loss": 2.0635,
+      "step": 20111
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3574666001673631e-05,
+      "loss": 2.0642,
+      "step": 20112
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3574106047384718e-05,
+      "loss": 2.1491,
+      "step": 20113
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.357354608024774e-05,
+      "loss": 2.4033,
+      "step": 20114
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3572986100264705e-05,
+      "loss": 1.3977,
+      "step": 20115
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3572426107437632e-05,
+      "loss": 2.0233,
+      "step": 20116
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.3571866101768534e-05,
+      "loss": 1.9518,
+      "step": 20117
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3571306083259418e-05,
+      "loss": 1.971,
+      "step": 20118
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3570746051912308e-05,
+      "loss": 1.8324,
+      "step": 20119
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3570186007729205e-05,
+      "loss": 1.6115,
+      "step": 20120
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.356962595071213e-05,
+      "loss": 1.8937,
+      "step": 20121
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.356906588086309e-05,
+      "loss": 1.8374,
+      "step": 20122
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.356850579818411e-05,
+      "loss": 1.9908,
+      "step": 20123
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3567945702677185e-05,
+      "loss": 1.7646,
+      "step": 20124
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3567385594344346e-05,
+      "loss": 1.7075,
+      "step": 20125
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3566825473187599e-05,
+      "loss": 2.0783,
+      "step": 20126
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3566265339208957e-05,
+      "loss": 1.5897,
+      "step": 20127
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3565705192410434e-05,
+      "loss": 1.9766,
+      "step": 20128
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3565145032794042e-05,
+      "loss": 2.0392,
+      "step": 20129
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3564584860361802e-05,
+      "loss": 1.9136,
+      "step": 20130
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3564024675115718e-05,
+      "loss": 2.2196,
+      "step": 20131
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.356346447705781e-05,
+      "loss": 2.153,
+      "step": 20132
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3562904266190091e-05,
+      "loss": 1.8726,
+      "step": 20133
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3562344042514572e-05,
+      "loss": 1.7436,
+      "step": 20134
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.356178380603327e-05,
+      "loss": 1.9717,
+      "step": 20135
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3561223556748195e-05,
+      "loss": 1.9642,
+      "step": 20136
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3560663294661369e-05,
+      "loss": 1.7299,
+      "step": 20137
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3560103019774795e-05,
+      "loss": 1.7732,
+      "step": 20138
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3559542732090494e-05,
+      "loss": 2.1986,
+      "step": 20139
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3558982431610479e-05,
+      "loss": 1.8565,
+      "step": 20140
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3558422118336765e-05,
+      "loss": 1.9459,
+      "step": 20141
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3557861792271364e-05,
+      "loss": 1.9463,
+      "step": 20142
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.355730145341629e-05,
+      "loss": 1.8696,
+      "step": 20143
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.355674110177356e-05,
+      "loss": 1.9356,
+      "step": 20144
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3556180737345185e-05,
+      "loss": 1.6768,
+      "step": 20145
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3555620360133184e-05,
+      "loss": 2.1168,
+      "step": 20146
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3555059970139568e-05,
+      "loss": 1.8299,
+      "step": 20147
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3554499567366352e-05,
+      "loss": 1.4716,
+      "step": 20148
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3553939151815552e-05,
+      "loss": 1.7036,
+      "step": 20149
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3553378723489178e-05,
+      "loss": 1.6947,
+      "step": 20150
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.355281828238925e-05,
+      "loss": 1.8803,
+      "step": 20151
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3552257828517779e-05,
+      "loss": 1.6465,
+      "step": 20152
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3551697361876781e-05,
+      "loss": 1.5751,
+      "step": 20153
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3551136882468274e-05,
+      "loss": 1.7515,
+      "step": 20154
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3550576390294267e-05,
+      "loss": 1.9558,
+      "step": 20155
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.355001588535678e-05,
+      "loss": 1.7125,
+      "step": 20156
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3549455367657824e-05,
+      "loss": 1.6756,
+      "step": 20157
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3548894837199415e-05,
+      "loss": 1.8979,
+      "step": 20158
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3548334293983567e-05,
+      "loss": 1.6059,
+      "step": 20159
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.35477737380123e-05,
+      "loss": 1.6981,
+      "step": 20160
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.354721316928762e-05,
+      "loss": 2.0533,
+      "step": 20161
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3546652587811554e-05,
+      "loss": 1.9507,
+      "step": 20162
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3546091993586108e-05,
+      "loss": 1.9978,
+      "step": 20163
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.35455313866133e-05,
+      "loss": 1.8791,
+      "step": 20164
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3544970766895144e-05,
+      "loss": 1.7324,
+      "step": 20165
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3544410134433657e-05,
+      "loss": 1.92,
+      "step": 20166
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3543849489230856e-05,
+      "loss": 2.0315,
+      "step": 20167
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3543288831288748e-05,
+      "loss": 1.8079,
+      "step": 20168
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.354272816060936e-05,
+      "loss": 2.0266,
+      "step": 20169
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3542167477194703e-05,
+      "loss": 1.8035,
+      "step": 20170
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3541606781046788e-05,
+      "loss": 1.8348,
+      "step": 20171
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3541046072167636e-05,
+      "loss": 1.9517,
+      "step": 20172
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3540485350559258e-05,
+      "loss": 1.8035,
+      "step": 20173
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3539924616223679e-05,
+      "loss": 1.6395,
+      "step": 20174
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3539363869162903e-05,
+      "loss": 1.7078,
+      "step": 20175
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.353880310937895e-05,
+      "loss": 2.0278,
+      "step": 20176
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3538242336873841e-05,
+      "loss": 1.8252,
+      "step": 20177
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3537681551649584e-05,
+      "loss": 2.17,
+      "step": 20178
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3537120753708201e-05,
+      "loss": 2.0442,
+      "step": 20179
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3536559943051702e-05,
+      "loss": 1.7556,
+      "step": 20180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3535999119682111e-05,
+      "loss": 1.8544,
+      "step": 20181
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5,
+      "learning_rate": 1.3535438283601437e-05,
+      "loss": 1.756,
+      "step": 20182
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3534877434811697e-05,
+      "loss": 1.4265,
+      "step": 20183
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3534316573314912e-05,
+      "loss": 2.2655,
+      "step": 20184
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3533755699113091e-05,
+      "loss": 1.8624,
+      "step": 20185
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3533194812208256e-05,
+      "loss": 1.7645,
+      "step": 20186
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3532633912602422e-05,
+      "loss": 1.8564,
+      "step": 20187
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3532073000297603e-05,
+      "loss": 1.7446,
+      "step": 20188
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3531512075295818e-05,
+      "loss": 1.7811,
+      "step": 20189
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.353095113759908e-05,
+      "loss": 2.3807,
+      "step": 20190
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3530390187209412e-05,
+      "loss": 2.0434,
+      "step": 20191
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.352982922412882e-05,
+      "loss": 1.5108,
+      "step": 20192
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3529268248359334e-05,
+      "loss": 2.0274,
+      "step": 20193
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3528707259902958e-05,
+      "loss": 1.8819,
+      "step": 20194
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3528146258761716e-05,
+      "loss": 2.0053,
+      "step": 20195
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3527585244937622e-05,
+      "loss": 1.8269,
+      "step": 20196
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3527024218432694e-05,
+      "loss": 1.9372,
+      "step": 20197
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3526463179248947e-05,
+      "loss": 1.7519,
+      "step": 20198
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3525902127388398e-05,
+      "loss": 1.6824,
+      "step": 20199
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3525341062853068e-05,
+      "loss": 1.8057,
+      "step": 20200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3524779985644967e-05,
+      "loss": 1.7769,
+      "step": 20201
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3524218895766116e-05,
+      "loss": 2.0892,
+      "step": 20202
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3523657793218535e-05,
+      "loss": 2.4064,
+      "step": 20203
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3523096678004235e-05,
+      "loss": 1.5535,
+      "step": 20204
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3522535550125238e-05,
+      "loss": 2.1624,
+      "step": 20205
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3521974409583553e-05,
+      "loss": 1.9085,
+      "step": 20206
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3521413256381208e-05,
+      "loss": 1.7658,
+      "step": 20207
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3520852090520214e-05,
+      "loss": 2.1245,
+      "step": 20208
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.352029091200259e-05,
+      "loss": 1.6471,
+      "step": 20209
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.351972972083035e-05,
+      "loss": 1.533,
+      "step": 20210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3519168517005516e-05,
+      "loss": 1.6964,
+      "step": 20211
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3518607300530104e-05,
+      "loss": 1.7877,
+      "step": 20212
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3518046071406129e-05,
+      "loss": 2.2671,
+      "step": 20213
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3517484829635614e-05,
+      "loss": 1.856,
+      "step": 20214
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3516923575220568e-05,
+      "loss": 1.7728,
+      "step": 20215
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3516362308163017e-05,
+      "loss": 1.8398,
+      "step": 20216
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3515801028464974e-05,
+      "loss": 1.9216,
+      "step": 20217
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3515239736128458e-05,
+      "loss": 1.7509,
+      "step": 20218
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3514678431155485e-05,
+      "loss": 1.763,
+      "step": 20219
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3514117113548073e-05,
+      "loss": 1.8502,
+      "step": 20220
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3513555783308247e-05,
+      "loss": 1.7663,
+      "step": 20221
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3512994440438014e-05,
+      "loss": 2.0513,
+      "step": 20222
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3512433084939398e-05,
+      "loss": 1.4985,
+      "step": 20223
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3511871716814416e-05,
+      "loss": 1.886,
+      "step": 20224
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3511310336065087e-05,
+      "loss": 1.7293,
+      "step": 20225
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3510748942693425e-05,
+      "loss": 1.6077,
+      "step": 20226
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3510187536701452e-05,
+      "loss": 1.7464,
+      "step": 20227
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3509626118091185e-05,
+      "loss": 1.908,
+      "step": 20228
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3509064686864642e-05,
+      "loss": 1.9105,
+      "step": 20229
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3508503243023843e-05,
+      "loss": 1.9659,
+      "step": 20230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3507941786570802e-05,
+      "loss": 1.8946,
+      "step": 20231
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3507380317507541e-05,
+      "loss": 1.9946,
+      "step": 20232
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3506818835836078e-05,
+      "loss": 1.9737,
+      "step": 20233
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.350625734155843e-05,
+      "loss": 1.8561,
+      "step": 20234
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3505695834676615e-05,
+      "loss": 1.8962,
+      "step": 20235
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3505134315192654e-05,
+      "loss": 1.9164,
+      "step": 20236
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3504572783108563e-05,
+      "loss": 1.6296,
+      "step": 20237
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3504011238426363e-05,
+      "loss": 1.7633,
+      "step": 20238
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3503449681148071e-05,
+      "loss": 2.1084,
+      "step": 20239
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3502888111275706e-05,
+      "loss": 1.935,
+      "step": 20240
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3502326528811286e-05,
+      "loss": 1.6907,
+      "step": 20241
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3501764933756829e-05,
+      "loss": 1.9478,
+      "step": 20242
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3501203326114356e-05,
+      "loss": 1.7192,
+      "step": 20243
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3500641705885889e-05,
+      "loss": 1.7775,
+      "step": 20244
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3500080073073436e-05,
+      "loss": 2.4171,
+      "step": 20245
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3499518427679028e-05,
+      "loss": 1.7216,
+      "step": 20246
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3498956769704679e-05,
+      "loss": 1.834,
+      "step": 20247
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3498395099152405e-05,
+      "loss": 2.3704,
+      "step": 20248
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3497833416024228e-05,
+      "loss": 2.1759,
+      "step": 20249
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3497271720322168e-05,
+      "loss": 1.8158,
+      "step": 20250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.3496710012048247e-05,
+      "loss": 1.1757,
+      "step": 20251
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3496148291204473e-05,
+      "loss": 1.8382,
+      "step": 20252
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.349558655779288e-05,
+      "loss": 1.7885,
+      "step": 20253
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3495024811815478e-05,
+      "loss": 1.9034,
+      "step": 20254
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3494463053274284e-05,
+      "loss": 1.7004,
+      "step": 20255
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3493901282171326e-05,
+      "loss": 1.5736,
+      "step": 20256
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3493339498508617e-05,
+      "loss": 2.0551,
+      "step": 20257
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.349277770228818e-05,
+      "loss": 1.5902,
+      "step": 20258
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3492215893512036e-05,
+      "loss": 1.4925,
+      "step": 20259
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3491654072182194e-05,
+      "loss": 1.8289,
+      "step": 20260
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3491092238300689e-05,
+      "loss": 2.2623,
+      "step": 20261
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.349053039186953e-05,
+      "loss": 1.914,
+      "step": 20262
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3489968532890738e-05,
+      "loss": 1.5947,
+      "step": 20263
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3489406661366337e-05,
+      "loss": 1.765,
+      "step": 20264
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3488844777298345e-05,
+      "loss": 1.6038,
+      "step": 20265
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3488282880688778e-05,
+      "loss": 1.6831,
+      "step": 20266
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.348772097153966e-05,
+      "loss": 1.9346,
+      "step": 20267
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3487159049853013e-05,
+      "loss": 1.9265,
+      "step": 20268
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.348659711563085e-05,
+      "loss": 1.7117,
+      "step": 20269
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3486035168875195e-05,
+      "loss": 1.967,
+      "step": 20270
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3485473209588072e-05,
+      "loss": 2.0214,
+      "step": 20271
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3484911237771495e-05,
+      "loss": 1.7675,
+      "step": 20272
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3484349253427485e-05,
+      "loss": 2.0271,
+      "step": 20273
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.348378725655806e-05,
+      "loss": 1.7162,
+      "step": 20274
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3483225247165252e-05,
+      "loss": 1.6286,
+      "step": 20275
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.3482663225251068e-05,
+      "loss": 1.9563,
+      "step": 20276
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3482101190817533e-05,
+      "loss": 1.817,
+      "step": 20277
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3481539143866669e-05,
+      "loss": 1.6934,
+      "step": 20278
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3480977084400495e-05,
+      "loss": 1.7998,
+      "step": 20279
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3480415012421033e-05,
+      "loss": 1.6722,
+      "step": 20280
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3479852927930301e-05,
+      "loss": 1.7043,
+      "step": 20281
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.347929083093032e-05,
+      "loss": 1.7677,
+      "step": 20282
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3478728721423111e-05,
+      "loss": 1.8222,
+      "step": 20283
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3478166599410696e-05,
+      "loss": 1.7482,
+      "step": 20284
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3477604464895096e-05,
+      "loss": 1.8928,
+      "step": 20285
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3477042317878329e-05,
+      "loss": 1.8654,
+      "step": 20286
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3476480158362416e-05,
+      "loss": 1.6705,
+      "step": 20287
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.347591798634938e-05,
+      "loss": 1.8838,
+      "step": 20288
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.347535580184124e-05,
+      "loss": 1.7753,
+      "step": 20289
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3474793604840018e-05,
+      "loss": 2.3072,
+      "step": 20290
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3474231395347736e-05,
+      "loss": 1.7057,
+      "step": 20291
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3473669173366412e-05,
+      "loss": 1.8414,
+      "step": 20292
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3473106938898071e-05,
+      "loss": 2.0037,
+      "step": 20293
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.347254469194473e-05,
+      "loss": 1.7348,
+      "step": 20294
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3471982432508414e-05,
+      "loss": 1.6118,
+      "step": 20295
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3471420160591142e-05,
+      "loss": 1.8287,
+      "step": 20296
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3470857876194932e-05,
+      "loss": 1.6681,
+      "step": 20297
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3470295579321815e-05,
+      "loss": 1.7795,
+      "step": 20298
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.34697332699738e-05,
+      "loss": 1.7033,
+      "step": 20299
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3469170948152917e-05,
+      "loss": 1.927,
+      "step": 20300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3468608613861184e-05,
+      "loss": 1.9895,
+      "step": 20301
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3468046267100625e-05,
+      "loss": 1.7647,
+      "step": 20302
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.346748390787326e-05,
+      "loss": 1.9946,
+      "step": 20303
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3466921536181107e-05,
+      "loss": 1.7112,
+      "step": 20304
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3466359152026197e-05,
+      "loss": 1.7615,
+      "step": 20305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.346579675541054e-05,
+      "loss": 2.0315,
+      "step": 20306
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3465234346336164e-05,
+      "loss": 1.6911,
+      "step": 20307
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3464671924805092e-05,
+      "loss": 1.9059,
+      "step": 20308
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3464109490819343e-05,
+      "loss": 1.8827,
+      "step": 20309
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.346354704438094e-05,
+      "loss": 2.1022,
+      "step": 20310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3462984585491903e-05,
+      "loss": 1.6977,
+      "step": 20311
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.346242211415426e-05,
+      "loss": 1.8691,
+      "step": 20312
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3461859630370022e-05,
+      "loss": 1.8197,
+      "step": 20313
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3461297134141221e-05,
+      "loss": 1.8528,
+      "step": 20314
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3460734625469875e-05,
+      "loss": 1.9106,
+      "step": 20315
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3460172104358007e-05,
+      "loss": 2.1273,
+      "step": 20316
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3459609570807636e-05,
+      "loss": 1.8083,
+      "step": 20317
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.345904702482079e-05,
+      "loss": 1.8263,
+      "step": 20318
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3458484466399488e-05,
+      "loss": 1.7814,
+      "step": 20319
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3457921895545747e-05,
+      "loss": 1.7458,
+      "step": 20320
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.34573593122616e-05,
+      "loss": 1.6454,
+      "step": 20321
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.345679671654906e-05,
+      "loss": 1.7193,
+      "step": 20322
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3456234108410158e-05,
+      "loss": 1.949,
+      "step": 20323
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3455671487846908e-05,
+      "loss": 2.0504,
+      "step": 20324
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3455108854861339e-05,
+      "loss": 1.8134,
+      "step": 20325
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.345454620945547e-05,
+      "loss": 2.0645,
+      "step": 20326
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3453983551631322e-05,
+      "loss": 1.8756,
+      "step": 20327
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3453420881390923e-05,
+      "loss": 2.0639,
+      "step": 20328
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3452858198736289e-05,
+      "loss": 1.8838,
+      "step": 20329
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3452295503669451e-05,
+      "loss": 1.8094,
+      "step": 20330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3451732796192425e-05,
+      "loss": 1.5653,
+      "step": 20331
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3451170076307234e-05,
+      "loss": 1.7458,
+      "step": 20332
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3450607344015905e-05,
+      "loss": 1.7291,
+      "step": 20333
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3450044599320458e-05,
+      "loss": 1.7524,
+      "step": 20334
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3449481842222918e-05,
+      "loss": 1.825,
+      "step": 20335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.3448919072725307e-05,
+      "loss": 1.9672,
+      "step": 20336
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3448356290829643e-05,
+      "loss": 2.1576,
+      "step": 20337
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3447793496537957e-05,
+      "loss": 1.6221,
+      "step": 20338
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3447230689852266e-05,
+      "loss": 2.0049,
+      "step": 20339
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3446667870774599e-05,
+      "loss": 1.5875,
+      "step": 20340
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3446105039306974e-05,
+      "loss": 2.0737,
+      "step": 20341
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.344554219545142e-05,
+      "loss": 2.0093,
+      "step": 20342
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3444979339209953e-05,
+      "loss": 1.9855,
+      "step": 20343
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.34444164705846e-05,
+      "loss": 1.6889,
+      "step": 20344
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.344385358957739e-05,
+      "loss": 1.9809,
+      "step": 20345
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3443290696190335e-05,
+      "loss": 1.9576,
+      "step": 20346
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3442727790425466e-05,
+      "loss": 1.9133,
+      "step": 20347
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3442164872284802e-05,
+      "loss": 1.6943,
+      "step": 20348
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3441601941770372e-05,
+      "loss": 1.9704,
+      "step": 20349
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.375,
+      "learning_rate": 1.3441038998884199e-05,
+      "loss": 2.4798,
+      "step": 20350
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.34404760436283e-05,
+      "loss": 1.8782,
+      "step": 20351
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3439913076004707e-05,
+      "loss": 1.8822,
+      "step": 20352
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.3439350096015437e-05,
+      "loss": 2.3879,
+      "step": 20353
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3438787103662518e-05,
+      "loss": 1.7459,
+      "step": 20354
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3438224098947973e-05,
+      "loss": 1.7979,
+      "step": 20355
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3437661081873827e-05,
+      "loss": 1.7391,
+      "step": 20356
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3437098052442102e-05,
+      "loss": 1.8889,
+      "step": 20357
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3436535010654819e-05,
+      "loss": 1.6316,
+      "step": 20358
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.343597195651401e-05,
+      "loss": 1.9001,
+      "step": 20359
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3435408890021689e-05,
+      "loss": 1.9575,
+      "step": 20360
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3434845811179888e-05,
+      "loss": 1.8511,
+      "step": 20361
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.343428271999063e-05,
+      "loss": 1.4029,
+      "step": 20362
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3433719616455937e-05,
+      "loss": 1.8505,
+      "step": 20363
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3433156500577834e-05,
+      "loss": 2.1167,
+      "step": 20364
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3432593372358347e-05,
+      "loss": 2.0005,
+      "step": 20365
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3432030231799495e-05,
+      "loss": 1.7748,
+      "step": 20366
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3431467078903306e-05,
+      "loss": 1.972,
+      "step": 20367
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3430903913671806e-05,
+      "loss": 2.1494,
+      "step": 20368
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3430340736107019e-05,
+      "loss": 1.7177,
+      "step": 20369
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.3429777546210966e-05,
+      "loss": 1.75,
+      "step": 20370
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3429214343985677e-05,
+      "loss": 1.9662,
+      "step": 20371
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3428651129433172e-05,
+      "loss": 2.342,
+      "step": 20372
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3428087902555478e-05,
+      "loss": 1.6087,
+      "step": 20373
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3427524663354614e-05,
+      "loss": 2.053,
+      "step": 20374
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3426961411832615e-05,
+      "loss": 1.7659,
+      "step": 20375
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3426398147991498e-05,
+      "loss": 1.6694,
+      "step": 20376
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3425834871833288e-05,
+      "loss": 1.8577,
+      "step": 20377
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3425271583360013e-05,
+      "loss": 1.959,
+      "step": 20378
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.34247082825737e-05,
+      "loss": 1.8165,
+      "step": 20379
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3424144969476367e-05,
+      "loss": 1.7448,
+      "step": 20380
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3423581644070042e-05,
+      "loss": 1.7671,
+      "step": 20381
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3423018306356754e-05,
+      "loss": 1.8758,
+      "step": 20382
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3422454956338522e-05,
+      "loss": 1.6142,
+      "step": 20383
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3421891594017375e-05,
+      "loss": 1.8538,
+      "step": 20384
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.3421328219395338e-05,
+      "loss": 1.9603,
+      "step": 20385
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3420764832474433e-05,
+      "loss": 1.9856,
+      "step": 20386
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 2.2148,
+      "step": 20387
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3419638021744129e-05,
+      "loss": 2.1029,
+      "step": 20388
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.341907459793878e-05,
+      "loss": 1.9001,
+      "step": 20389
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3418511161842664e-05,
+      "loss": 1.7746,
+      "step": 20390
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.341794771345781e-05,
+      "loss": 1.7268,
+      "step": 20391
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3417384252786244e-05,
+      "loss": 2.0293,
+      "step": 20392
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3416820779829988e-05,
+      "loss": 1.8732,
+      "step": 20393
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.341625729459107e-05,
+      "loss": 1.7209,
+      "step": 20394
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3415693797071514e-05,
+      "loss": 1.8596,
+      "step": 20395
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3415130287273349e-05,
+      "loss": 1.872,
+      "step": 20396
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3414566765198593e-05,
+      "loss": 1.6675,
+      "step": 20397
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3414003230849283e-05,
+      "loss": 2.0249,
+      "step": 20398
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3413439684227436e-05,
+      "loss": 1.9845,
+      "step": 20399
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3412876125335082e-05,
+      "loss": 1.9496,
+      "step": 20400
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3412312554174243e-05,
+      "loss": 2.2005,
+      "step": 20401
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.125,
+      "learning_rate": 1.341174897074695e-05,
+      "loss": 1.6232,
+      "step": 20402
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3411185375055227e-05,
+      "loss": 1.7711,
+      "step": 20403
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3410621767101093e-05,
+      "loss": 1.94,
+      "step": 20404
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3410058146886586e-05,
+      "loss": 2.1032,
+      "step": 20405
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3409494514413725e-05,
+      "loss": 2.1251,
+      "step": 20406
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3408930869684538e-05,
+      "loss": 2.0725,
+      "step": 20407
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.340836721270105e-05,
+      "loss": 2.0445,
+      "step": 20408
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.340780354346529e-05,
+      "loss": 1.8546,
+      "step": 20409
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.340723986197928e-05,
+      "loss": 1.9118,
+      "step": 20410
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3406676168245047e-05,
+      "loss": 1.8709,
+      "step": 20411
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3406112462264624e-05,
+      "loss": 1.9356,
+      "step": 20412
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3405548744040026e-05,
+      "loss": 1.9112,
+      "step": 20413
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.340498501357329e-05,
+      "loss": 1.6978,
+      "step": 20414
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3404421270866436e-05,
+      "loss": 1.8614,
+      "step": 20415
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3403857515921495e-05,
+      "loss": 2.0061,
+      "step": 20416
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.340329374874049e-05,
+      "loss": 1.6167,
+      "step": 20417
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3402729969325447e-05,
+      "loss": 1.881,
+      "step": 20418
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3402166177678397e-05,
+      "loss": 1.9294,
+      "step": 20419
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3401602373801365e-05,
+      "loss": 2.186,
+      "step": 20420
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3401038557696374e-05,
+      "loss": 1.8525,
+      "step": 20421
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3400474729365457e-05,
+      "loss": 1.9045,
+      "step": 20422
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3399910888810632e-05,
+      "loss": 1.7231,
+      "step": 20423
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3399347036033936e-05,
+      "loss": 2.0174,
+      "step": 20424
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3398783171037388e-05,
+      "loss": 1.7314,
+      "step": 20425
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3398219293823022e-05,
+      "loss": 2.022,
+      "step": 20426
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3397655404392859e-05,
+      "loss": 1.4244,
+      "step": 20427
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.339709150274893e-05,
+      "loss": 1.6612,
+      "step": 20428
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.339652758889326e-05,
+      "loss": 1.867,
+      "step": 20429
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3395963662827876e-05,
+      "loss": 1.8867,
+      "step": 20430
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3395399724554805e-05,
+      "loss": 1.9826,
+      "step": 20431
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3394835774076077e-05,
+      "loss": 1.8877,
+      "step": 20432
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3394271811393715e-05,
+      "loss": 2.0241,
+      "step": 20433
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.339370783650975e-05,
+      "loss": 1.7639,
+      "step": 20434
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.3393143849426205e-05,
+      "loss": 1.8934,
+      "step": 20435
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3392579850145116e-05,
+      "loss": 1.7317,
+      "step": 20436
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.33920158386685e-05,
+      "loss": 2.0447,
+      "step": 20437
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.3391451814998391e-05,
+      "loss": 1.7816,
+      "step": 20438
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3390887779136814e-05,
+      "loss": 1.6609,
+      "step": 20439
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.33903237310858e-05,
+      "loss": 1.9843,
+      "step": 20440
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3389759670847371e-05,
+      "loss": 1.8401,
+      "step": 20441
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3389195598423556e-05,
+      "loss": 2.4027,
+      "step": 20442
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.338863151381639e-05,
+      "loss": 1.7188,
+      "step": 20443
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.338806741702789e-05,
+      "loss": 2.0635,
+      "step": 20444
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3387503308060091e-05,
+      "loss": 1.7023,
+      "step": 20445
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.338693918691502e-05,
+      "loss": 1.6213,
+      "step": 20446
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3386375053594703e-05,
+      "loss": 2.1928,
+      "step": 20447
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3385810908101167e-05,
+      "loss": 1.7588,
+      "step": 20448
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3385246750436442e-05,
+      "loss": 1.9369,
+      "step": 20449
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3384682580602557e-05,
+      "loss": 2.039,
+      "step": 20450
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3384118398601535e-05,
+      "loss": 2.1548,
+      "step": 20451
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.338355420443541e-05,
+      "loss": 1.8307,
+      "step": 20452
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3382989998106209e-05,
+      "loss": 1.8456,
+      "step": 20453
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.3382425779615958e-05,
+      "loss": 1.7013,
+      "step": 20454
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3381861548966686e-05,
+      "loss": 1.9174,
+      "step": 20455
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3381297306160423e-05,
+      "loss": 2.0213,
+      "step": 20456
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3380733051199194e-05,
+      "loss": 2.0466,
+      "step": 20457
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3380168784085028e-05,
+      "loss": 1.87,
+      "step": 20458
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3379604504819958e-05,
+      "loss": 1.7372,
+      "step": 20459
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3379040213406007e-05,
+      "loss": 1.588,
+      "step": 20460
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.3378475909845205e-05,
+      "loss": 2.1235,
+      "step": 20461
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3377911594139583e-05,
+      "loss": 1.6393,
+      "step": 20462
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3377347266291167e-05,
+      "loss": 2.0202,
+      "step": 20463
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3376782926301985e-05,
+      "loss": 1.9391,
+      "step": 20464
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3376218574174067e-05,
+      "loss": 1.832,
+      "step": 20465
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.3375654209909446e-05,
+      "loss": 1.4629,
+      "step": 20466
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.3375089833510139e-05,
+      "loss": 1.2212,
+      "step": 20467
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3374525444978188e-05,
+      "loss": 1.9242,
+      "step": 20468
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3373961044315615e-05,
+      "loss": 2.0872,
+      "step": 20469
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3373396631524449e-05,
+      "loss": 1.6878,
+      "step": 20470
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.337283220660672e-05,
+      "loss": 2.1884,
+      "step": 20471
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3372267769564456e-05,
+      "loss": 2.1913,
+      "step": 20472
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.337170332039969e-05,
+      "loss": 2.0412,
+      "step": 20473
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3371138859114445e-05,
+      "loss": 2.2075,
+      "step": 20474
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3370574385710753e-05,
+      "loss": 1.8749,
+      "step": 20475
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3370009900190647e-05,
+      "loss": 1.7878,
+      "step": 20476
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.336944540255615e-05,
+      "loss": 1.8274,
+      "step": 20477
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.336888089280929e-05,
+      "loss": 2.0411,
+      "step": 20478
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3368316370952103e-05,
+      "loss": 1.6248,
+      "step": 20479
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3367751836986618e-05,
+      "loss": 1.9692,
+      "step": 20480
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3367187290914857e-05,
+      "loss": 1.8452,
+      "step": 20481
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3366622732738857e-05,
+      "loss": 1.6859,
+      "step": 20482
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3366058162460642e-05,
+      "loss": 1.8253,
+      "step": 20483
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3365493580082247e-05,
+      "loss": 1.7987,
+      "step": 20484
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3364928985605696e-05,
+      "loss": 1.6674,
+      "step": 20485
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3364364379033024e-05,
+      "loss": 1.9302,
+      "step": 20486
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3363799760366253e-05,
+      "loss": 1.4897,
+      "step": 20487
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3363235129607422e-05,
+      "loss": 1.9631,
+      "step": 20488
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.3362670486758553e-05,
+      "loss": 2.0515,
+      "step": 20489
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.336210583182168e-05,
+      "loss": 2.1467,
+      "step": 20490
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3361541164798833e-05,
+      "loss": 1.8228,
+      "step": 20491
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3360976485692038e-05,
+      "loss": 1.6905,
+      "step": 20492
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.336041179450333e-05,
+      "loss": 2.1272,
+      "step": 20493
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3359847091234734e-05,
+      "loss": 2.3376,
+      "step": 20494
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.335928237588828e-05,
+      "loss": 1.5268,
+      "step": 20495
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3358717648466006e-05,
+      "loss": 1.8915,
+      "step": 20496
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3358152908969935e-05,
+      "loss": 1.6739,
+      "step": 20497
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3357588157402094e-05,
+      "loss": 2.1192,
+      "step": 20498
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.3357023393764523e-05,
+      "loss": 2.6853,
+      "step": 20499
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3356458618059245e-05,
+      "loss": 1.8682,
+      "step": 20500
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3355893830288291e-05,
+      "loss": 1.7189,
+      "step": 20501
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3355329030453692e-05,
+      "loss": 1.8807,
+      "step": 20502
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3354764218557482e-05,
+      "loss": 1.8916,
+      "step": 20503
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3354199394601685e-05,
+      "loss": 1.6237,
+      "step": 20504
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3353634558588335e-05,
+      "loss": 1.9423,
+      "step": 20505
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3353069710519464e-05,
+      "loss": 1.7883,
+      "step": 20506
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3352504850397097e-05,
+      "loss": 2.2977,
+      "step": 20507
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.335193997822327e-05,
+      "loss": 1.9707,
+      "step": 20508
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3351375094000007e-05,
+      "loss": 2.0266,
+      "step": 20509
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3350810197729352e-05,
+      "loss": 1.5142,
+      "step": 20510
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3350245289413319e-05,
+      "loss": 1.7668,
+      "step": 20511
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3349680369053948e-05,
+      "loss": 1.7903,
+      "step": 20512
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3349115436653272e-05,
+      "loss": 1.7022,
+      "step": 20513
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3348550492213315e-05,
+      "loss": 1.5415,
+      "step": 20514
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3347985535736112e-05,
+      "loss": 1.74,
+      "step": 20515
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.334742056722369e-05,
+      "loss": 1.9003,
+      "step": 20516
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3346855586678087e-05,
+      "loss": 2.0886,
+      "step": 20517
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3346290594101327e-05,
+      "loss": 1.7451,
+      "step": 20518
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.3345725589495443e-05,
+      "loss": 1.885,
+      "step": 20519
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.334516057286247e-05,
+      "loss": 1.8121,
+      "step": 20520
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3344595544204432e-05,
+      "loss": 2.1622,
+      "step": 20521
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3344030503523367e-05,
+      "loss": 1.9168,
+      "step": 20522
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3343465450821302e-05,
+      "loss": 2.0195,
+      "step": 20523
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3342900386100268e-05,
+      "loss": 1.6819,
+      "step": 20524
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3342335309362299e-05,
+      "loss": 2.0395,
+      "step": 20525
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3341770220609424e-05,
+      "loss": 1.8855,
+      "step": 20526
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3341205119843679e-05,
+      "loss": 2.2857,
+      "step": 20527
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3340640007067087e-05,
+      "loss": 1.9403,
+      "step": 20528
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3340074882281688e-05,
+      "loss": 1.8954,
+      "step": 20529
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.3339509745489507e-05,
+      "loss": 2.1425,
+      "step": 20530
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.333894459669258e-05,
+      "loss": 1.8744,
+      "step": 20531
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3338379435892936e-05,
+      "loss": 2.1646,
+      "step": 20532
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.333781426309261e-05,
+      "loss": 1.7621,
+      "step": 20533
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3337249078293628e-05,
+      "loss": 1.7225,
+      "step": 20534
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3336683881498024e-05,
+      "loss": 2.2156,
+      "step": 20535
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3336118672707832e-05,
+      "loss": 1.8121,
+      "step": 20536
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3335553451925082e-05,
+      "loss": 1.9471,
+      "step": 20537
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3334988219151807e-05,
+      "loss": 2.3465,
+      "step": 20538
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.3334422974390037e-05,
+      "loss": 1.8938,
+      "step": 20539
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3333857717641807e-05,
+      "loss": 1.9453,
+      "step": 20540
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.3333292448909145e-05,
+      "loss": 1.8158,
+      "step": 20541
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3332727168194085e-05,
+      "loss": 2.1337,
+      "step": 20542
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3332161875498661e-05,
+      "loss": 1.9218,
+      "step": 20543
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.33315965708249e-05,
+      "loss": 1.6921,
+      "step": 20544
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.333103125417484e-05,
+      "loss": 1.8498,
+      "step": 20545
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.333046592555051e-05,
+      "loss": 1.6166,
+      "step": 20546
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.332990058495394e-05,
+      "loss": 1.5752,
+      "step": 20547
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3329335232387169e-05,
+      "loss": 1.7534,
+      "step": 20548
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.332876986785222e-05,
+      "loss": 1.8951,
+      "step": 20549
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3328204491351136e-05,
+      "loss": 1.832,
+      "step": 20550
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3327639102885939e-05,
+      "loss": 1.971,
+      "step": 20551
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3327073702458671e-05,
+      "loss": 1.8644,
+      "step": 20552
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3326508290071356e-05,
+      "loss": 1.7734,
+      "step": 20553
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3325942865726033e-05,
+      "loss": 1.6879,
+      "step": 20554
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3325377429424733e-05,
+      "loss": 2.2091,
+      "step": 20555
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.3324811981169483e-05,
+      "loss": 1.9937,
+      "step": 20556
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3324246520962325e-05,
+      "loss": 1.8993,
+      "step": 20557
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3323681048805282e-05,
+      "loss": 2.0423,
+      "step": 20558
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3323115564700395e-05,
+      "loss": 1.6452,
+      "step": 20559
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3322550068649693e-05,
+      "loss": 1.3747,
+      "step": 20560
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3321984560655211e-05,
+      "loss": 1.4997,
+      "step": 20561
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3321419040718977e-05,
+      "loss": 1.8819,
+      "step": 20562
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3320853508843028e-05,
+      "loss": 1.7887,
+      "step": 20563
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3320287965029397e-05,
+      "loss": 2.0157,
+      "step": 20564
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3319722409280114e-05,
+      "loss": 1.7907,
+      "step": 20565
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3319156841597216e-05,
+      "loss": 1.9572,
+      "step": 20566
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3318591261982734e-05,
+      "loss": 2.099,
+      "step": 20567
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.33180256704387e-05,
+      "loss": 2.1386,
+      "step": 20568
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.331746006696715e-05,
+      "loss": 1.7507,
+      "step": 20569
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3316894451570115e-05,
+      "loss": 2.0242,
+      "step": 20570
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.331632882424963e-05,
+      "loss": 2.1608,
+      "step": 20571
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3315763185007723e-05,
+      "loss": 1.818,
+      "step": 20572
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3315197533846433e-05,
+      "loss": 1.8383,
+      "step": 20573
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3314631870767794e-05,
+      "loss": 1.8257,
+      "step": 20574
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3314066195773838e-05,
+      "loss": 1.6902,
+      "step": 20575
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3313500508866595e-05,
+      "loss": 1.9044,
+      "step": 20576
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3312934810048103e-05,
+      "loss": 1.7352,
+      "step": 20577
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3312369099320395e-05,
+      "loss": 1.9729,
+      "step": 20578
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.33118033766855e-05,
+      "loss": 1.5825,
+      "step": 20579
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3311237642145459e-05,
+      "loss": 1.9699,
+      "step": 20580
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3310671895702298e-05,
+      "loss": 1.9002,
+      "step": 20581
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3310106137358056e-05,
+      "loss": 2.1064,
+      "step": 20582
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3309540367114767e-05,
+      "loss": 2.0508,
+      "step": 20583
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3308974584974462e-05,
+      "loss": 1.9479,
+      "step": 20584
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3308408790939174e-05,
+      "loss": 1.8472,
+      "step": 20585
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.330784298501094e-05,
+      "loss": 2.0101,
+      "step": 20586
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3307277167191794e-05,
+      "loss": 1.8546,
+      "step": 20587
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3306711337483767e-05,
+      "loss": 1.8616,
+      "step": 20588
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3306145495888894e-05,
+      "loss": 1.8627,
+      "step": 20589
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3305579642409214e-05,
+      "loss": 1.851,
+      "step": 20590
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3305013777046751e-05,
+      "loss": 2.324,
+      "step": 20591
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.330444789980355e-05,
+      "loss": 1.8008,
+      "step": 20592
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3303882010681637e-05,
+      "loss": 1.9201,
+      "step": 20593
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3303316109683053e-05,
+      "loss": 1.9011,
+      "step": 20594
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3302750196809825e-05,
+      "loss": 1.7941,
+      "step": 20595
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3302184272063993e-05,
+      "loss": 1.768,
+      "step": 20596
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3301618335447592e-05,
+      "loss": 2.0208,
+      "step": 20597
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3301052386962649e-05,
+      "loss": 2.0202,
+      "step": 20598
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3300486426611206e-05,
+      "loss": 1.9805,
+      "step": 20599
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3299920454395296e-05,
+      "loss": 1.9027,
+      "step": 20600
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3299354470316952e-05,
+      "loss": 1.8714,
+      "step": 20601
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3298788474378208e-05,
+      "loss": 2.1318,
+      "step": 20602
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3298222466581097e-05,
+      "loss": 1.8692,
+      "step": 20603
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.329765644692766e-05,
+      "loss": 2.1384,
+      "step": 20604
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3297090415419924e-05,
+      "loss": 1.6027,
+      "step": 20605
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3296524372059931e-05,
+      "loss": 1.9693,
+      "step": 20606
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3295958316849712e-05,
+      "loss": 2.0656,
+      "step": 20607
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3295392249791303e-05,
+      "loss": 1.9638,
+      "step": 20608
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3294826170886738e-05,
+      "loss": 2.0794,
+      "step": 20609
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3294260080138051e-05,
+      "loss": 1.5388,
+      "step": 20610
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3293693977547279e-05,
+      "loss": 2.0285,
+      "step": 20611
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3293127863116455e-05,
+      "loss": 1.7793,
+      "step": 20612
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3292561736847615e-05,
+      "loss": 2.1378,
+      "step": 20613
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3291995598742795e-05,
+      "loss": 1.9795,
+      "step": 20614
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.329142944880403e-05,
+      "loss": 1.9102,
+      "step": 20615
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3290863287033356e-05,
+      "loss": 1.7349,
+      "step": 20616
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3290297113432805e-05,
+      "loss": 1.9654,
+      "step": 20617
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3289730928004413e-05,
+      "loss": 1.8086,
+      "step": 20618
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.3289164730750214e-05,
+      "loss": 2.1589,
+      "step": 20619
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3288598521672249e-05,
+      "loss": 1.5042,
+      "step": 20620
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3288032300772551e-05,
+      "loss": 1.8449,
+      "step": 20621
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3287466068053153e-05,
+      "loss": 1.9189,
+      "step": 20622
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3286899823516093e-05,
+      "loss": 1.8222,
+      "step": 20623
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3286333567163406e-05,
+      "loss": 1.6689,
+      "step": 20624
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3285767298997123e-05,
+      "loss": 1.711,
+      "step": 20625
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3285201019019285e-05,
+      "loss": 1.9483,
+      "step": 20626
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.328463472723193e-05,
+      "loss": 1.892,
+      "step": 20627
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3284068423637087e-05,
+      "loss": 1.9236,
+      "step": 20628
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3283502108236794e-05,
+      "loss": 1.7485,
+      "step": 20629
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3282935781033089e-05,
+      "loss": 1.7834,
+      "step": 20630
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3282369442028006e-05,
+      "loss": 2.1296,
+      "step": 20631
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3281803091223582e-05,
+      "loss": 1.8434,
+      "step": 20632
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3281236728621849e-05,
+      "loss": 2.0403,
+      "step": 20633
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.328067035422485e-05,
+      "loss": 1.7675,
+      "step": 20634
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3280103968034612e-05,
+      "loss": 1.7854,
+      "step": 20635
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.375,
+      "learning_rate": 1.327953757005318e-05,
+      "loss": 1.9502,
+      "step": 20636
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3278971160282586e-05,
+      "loss": 1.8307,
+      "step": 20637
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3278404738724864e-05,
+      "loss": 1.8063,
+      "step": 20638
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3277838305382053e-05,
+      "loss": 2.1916,
+      "step": 20639
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3277271860256188e-05,
+      "loss": 1.6777,
+      "step": 20640
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3276705403349308e-05,
+      "loss": 1.7218,
+      "step": 20641
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3276138934663444e-05,
+      "loss": 1.7405,
+      "step": 20642
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3275572454200636e-05,
+      "loss": 1.9185,
+      "step": 20643
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.327500596196292e-05,
+      "loss": 1.7621,
+      "step": 20644
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3274439457952335e-05,
+      "loss": 2.0266,
+      "step": 20645
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.327387294217091e-05,
+      "loss": 1.8715,
+      "step": 20646
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3273306414620688e-05,
+      "loss": 1.5109,
+      "step": 20647
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3272739875303706e-05,
+      "loss": 2.0324,
+      "step": 20648
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3272173324221994e-05,
+      "loss": 1.843,
+      "step": 20649
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3271606761377594e-05,
+      "loss": 1.8716,
+      "step": 20650
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3271040186772544e-05,
+      "loss": 1.7111,
+      "step": 20651
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3270473600408875e-05,
+      "loss": 1.9872,
+      "step": 20652
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.326990700228863e-05,
+      "loss": 2.0452,
+      "step": 20653
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3269340392413838e-05,
+      "loss": 2.0533,
+      "step": 20654
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3268773770786543e-05,
+      "loss": 1.7672,
+      "step": 20655
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3268207137408777e-05,
+      "loss": 1.6564,
+      "step": 20656
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3267640492282583e-05,
+      "loss": 1.6768,
+      "step": 20657
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3267073835409993e-05,
+      "loss": 1.9741,
+      "step": 20658
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3266507166793046e-05,
+      "loss": 2.0572,
+      "step": 20659
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3265940486433777e-05,
+      "loss": 1.7197,
+      "step": 20660
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.3265373794334224e-05,
+      "loss": 2.3157,
+      "step": 20661
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3264807090496425e-05,
+      "loss": 1.8035,
+      "step": 20662
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3264240374922417e-05,
+      "loss": 1.829,
+      "step": 20663
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3263673647614236e-05,
+      "loss": 2.0682,
+      "step": 20664
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3263106908573921e-05,
+      "loss": 2.0005,
+      "step": 20665
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3262540157803507e-05,
+      "loss": 1.8142,
+      "step": 20666
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3261973395305034e-05,
+      "loss": 1.7961,
+      "step": 20667
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3261406621080538e-05,
+      "loss": 1.7483,
+      "step": 20668
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3260839835132055e-05,
+      "loss": 1.8407,
+      "step": 20669
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3260273037461625e-05,
+      "loss": 1.875,
+      "step": 20670
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3259706228071286e-05,
+      "loss": 1.5806,
+      "step": 20671
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3259139406963073e-05,
+      "loss": 1.527,
+      "step": 20672
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3258572574139024e-05,
+      "loss": 2.1167,
+      "step": 20673
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3258005729601178e-05,
+      "loss": 2.0113,
+      "step": 20674
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3257438873351569e-05,
+      "loss": 1.8152,
+      "step": 20675
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.325687200539224e-05,
+      "loss": 2.0604,
+      "step": 20676
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3256305125725227e-05,
+      "loss": 2.1174,
+      "step": 20677
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3255738234352566e-05,
+      "loss": 1.8471,
+      "step": 20678
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3255171331276297e-05,
+      "loss": 2.1191,
+      "step": 20679
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3254604416498456e-05,
+      "loss": 1.8989,
+      "step": 20680
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3254037490021083e-05,
+      "loss": 1.835,
+      "step": 20681
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.3253470551846213e-05,
+      "loss": 1.8495,
+      "step": 20682
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3252903601975887e-05,
+      "loss": 1.8903,
+      "step": 20683
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3252336640412142e-05,
+      "loss": 1.8376,
+      "step": 20684
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3251769667157015e-05,
+      "loss": 2.0386,
+      "step": 20685
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3251202682212546e-05,
+      "loss": 1.7665,
+      "step": 20686
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.325063568558077e-05,
+      "loss": 1.7303,
+      "step": 20687
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3250068677263732e-05,
+      "loss": 1.7068,
+      "step": 20688
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3249501657263462e-05,
+      "loss": 1.8191,
+      "step": 20689
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3248934625582001e-05,
+      "loss": 1.8012,
+      "step": 20690
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.3248367582221393e-05,
+      "loss": 2.3931,
+      "step": 20691
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3247800527183668e-05,
+      "loss": 1.8971,
+      "step": 20692
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.324723346047087e-05,
+      "loss": 1.9093,
+      "step": 20693
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3246666382085031e-05,
+      "loss": 1.8274,
+      "step": 20694
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3246099292028202e-05,
+      "loss": 1.8876,
+      "step": 20695
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3245532190302407e-05,
+      "loss": 2.0298,
+      "step": 20696
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3244965076909697e-05,
+      "loss": 1.9023,
+      "step": 20697
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3244397951852102e-05,
+      "loss": 1.7979,
+      "step": 20698
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3243830815131664e-05,
+      "loss": 1.9144,
+      "step": 20699
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.324326366675042e-05,
+      "loss": 2.0603,
+      "step": 20700
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3242696506710413e-05,
+      "loss": 1.5028,
+      "step": 20701
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3242129335013676e-05,
+      "loss": 2.2417,
+      "step": 20702
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.324156215166225e-05,
+      "loss": 1.6947,
+      "step": 20703
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3240994956658177e-05,
+      "loss": 1.6593,
+      "step": 20704
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3240427750003495e-05,
+      "loss": 1.9977,
+      "step": 20705
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3239860531700239e-05,
+      "loss": 2.0715,
+      "step": 20706
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3239293301750454e-05,
+      "loss": 1.9551,
+      "step": 20707
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.323872606015617e-05,
+      "loss": 1.8029,
+      "step": 20708
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3238158806919437e-05,
+      "loss": 2.2184,
+      "step": 20709
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3237591542042286e-05,
+      "loss": 1.9788,
+      "step": 20710
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3237024265526761e-05,
+      "loss": 1.5917,
+      "step": 20711
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3236456977374896e-05,
+      "loss": 1.9649,
+      "step": 20712
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3235889677588736e-05,
+      "loss": 1.9825,
+      "step": 20713
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3235322366170317e-05,
+      "loss": 1.6897,
+      "step": 20714
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3234755043121681e-05,
+      "loss": 2.303,
+      "step": 20715
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.3234187708444866e-05,
+      "loss": 1.9831,
+      "step": 20716
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3233620362141906e-05,
+      "loss": 1.7459,
+      "step": 20717
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3233053004214851e-05,
+      "loss": 1.9077,
+      "step": 20718
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3232485634665732e-05,
+      "loss": 1.7278,
+      "step": 20719
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.3231918253496594e-05,
+      "loss": 1.6491,
+      "step": 20720
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3231350860709472e-05,
+      "loss": 1.9683,
+      "step": 20721
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3230783456306409e-05,
+      "loss": 2.1202,
+      "step": 20722
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3230216040289444e-05,
+      "loss": 2.0011,
+      "step": 20723
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3229648612660612e-05,
+      "loss": 1.7353,
+      "step": 20724
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3229081173421964e-05,
+      "loss": 1.9262,
+      "step": 20725
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.322851372257553e-05,
+      "loss": 1.8375,
+      "step": 20726
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.322794626012335e-05,
+      "loss": 1.9253,
+      "step": 20727
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.322737878606747e-05,
+      "loss": 2.0931,
+      "step": 20728
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3226811300409925e-05,
+      "loss": 1.9937,
+      "step": 20729
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3226243803152756e-05,
+      "loss": 2.2406,
+      "step": 20730
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3225676294298005e-05,
+      "loss": 1.7273,
+      "step": 20731
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.322510877384771e-05,
+      "loss": 2.0283,
+      "step": 20732
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.322454124180391e-05,
+      "loss": 1.8808,
+      "step": 20733
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3223973698168649e-05,
+      "loss": 2.0194,
+      "step": 20734
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3223406142943965e-05,
+      "loss": 1.7652,
+      "step": 20735
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3222838576131896e-05,
+      "loss": 1.9702,
+      "step": 20736
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3222270997734486e-05,
+      "loss": 1.8357,
+      "step": 20737
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3221703407753774e-05,
+      "loss": 2.3921,
+      "step": 20738
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.32211358061918e-05,
+      "loss": 1.9621,
+      "step": 20739
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3220568193050602e-05,
+      "loss": 2.0683,
+      "step": 20740
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3220000568332227e-05,
+      "loss": 1.5954,
+      "step": 20741
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3219432932038712e-05,
+      "loss": 1.8228,
+      "step": 20742
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3218865284172094e-05,
+      "loss": 1.9764,
+      "step": 20743
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3218297624734416e-05,
+      "loss": 1.8783,
+      "step": 20744
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.3217729953727721e-05,
+      "loss": 2.3453,
+      "step": 20745
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3217162271154048e-05,
+      "loss": 1.6647,
+      "step": 20746
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3216594577015435e-05,
+      "loss": 1.9033,
+      "step": 20747
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3216026871313929e-05,
+      "loss": 2.0519,
+      "step": 20748
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3215459154051566e-05,
+      "loss": 1.9043,
+      "step": 20749
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3214891425230385e-05,
+      "loss": 1.8812,
+      "step": 20750
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3214323684852431e-05,
+      "loss": 2.0526,
+      "step": 20751
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3213755932919746e-05,
+      "loss": 1.9842,
+      "step": 20752
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.3213188169434365e-05,
+      "loss": 1.8598,
+      "step": 20753
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3212620394398333e-05,
+      "loss": 1.8758,
+      "step": 20754
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3212052607813693e-05,
+      "loss": 1.9583,
+      "step": 20755
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3211484809682482e-05,
+      "loss": 1.9185,
+      "step": 20756
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.321091700000674e-05,
+      "loss": 1.7741,
+      "step": 20757
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3210349178788518e-05,
+      "loss": 2.1656,
+      "step": 20758
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3209781346029842e-05,
+      "loss": 1.7769,
+      "step": 20759
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3209213501732764e-05,
+      "loss": 2.0536,
+      "step": 20760
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3208645645899323e-05,
+      "loss": 1.9698,
+      "step": 20761
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.320807777853156e-05,
+      "loss": 1.9875,
+      "step": 20762
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.3207509899631515e-05,
+      "loss": 1.9386,
+      "step": 20763
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.320694200920123e-05,
+      "loss": 2.0178,
+      "step": 20764
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.320637410724275e-05,
+      "loss": 2.726,
+      "step": 20765
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3205806193758109e-05,
+      "loss": 2.0344,
+      "step": 20766
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3205238268749354e-05,
+      "loss": 1.5693,
+      "step": 20767
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3204670332218528e-05,
+      "loss": 1.9373,
+      "step": 20768
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3204102384167668e-05,
+      "loss": 2.0764,
+      "step": 20769
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3203534424598819e-05,
+      "loss": 2.005,
+      "step": 20770
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.320296645351402e-05,
+      "loss": 1.6467,
+      "step": 20771
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3202398470915316e-05,
+      "loss": 1.8442,
+      "step": 20772
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3201830476804744e-05,
+      "loss": 1.7162,
+      "step": 20773
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3201262471184348e-05,
+      "loss": 2.0973,
+      "step": 20774
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3200694454056176e-05,
+      "loss": 2.0641,
+      "step": 20775
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3200126425422259e-05,
+      "loss": 2.0668,
+      "step": 20776
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3199558385284647e-05,
+      "loss": 1.9996,
+      "step": 20777
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3198990333645378e-05,
+      "loss": 1.9675,
+      "step": 20778
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3198422270506496e-05,
+      "loss": 1.9056,
+      "step": 20779
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3197854195870042e-05,
+      "loss": 1.9232,
+      "step": 20780
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3197286109738058e-05,
+      "loss": 1.6888,
+      "step": 20781
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.3196718012112588e-05,
+      "loss": 1.7751,
+      "step": 20782
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3196149902995672e-05,
+      "loss": 1.982,
+      "step": 20783
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3195581782389354e-05,
+      "loss": 1.7342,
+      "step": 20784
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3195013650295673e-05,
+      "loss": 1.8814,
+      "step": 20785
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3194445506716676e-05,
+      "loss": 1.7694,
+      "step": 20786
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3193877351654401e-05,
+      "loss": 2.1169,
+      "step": 20787
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3193309185110891e-05,
+      "loss": 1.8196,
+      "step": 20788
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.3192741007088192e-05,
+      "loss": 2.0853,
+      "step": 20789
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3192172817588344e-05,
+      "loss": 1.6246,
+      "step": 20790
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3191604616613389e-05,
+      "loss": 1.809,
+      "step": 20791
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3191036404165372e-05,
+      "loss": 1.7866,
+      "step": 20792
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.3190468180246332e-05,
+      "loss": 1.737,
+      "step": 20793
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3189899944858313e-05,
+      "loss": 1.9262,
+      "step": 20794
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.318933169800336e-05,
+      "loss": 1.97,
+      "step": 20795
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3188763439683512e-05,
+      "loss": 2.0337,
+      "step": 20796
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3188195169900813e-05,
+      "loss": 1.806,
+      "step": 20797
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3187626888657308e-05,
+      "loss": 1.9372,
+      "step": 20798
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3187058595955037e-05,
+      "loss": 1.9082,
+      "step": 20799
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3186490291796046e-05,
+      "loss": 2.024,
+      "step": 20800
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3185921976182372e-05,
+      "loss": 2.018,
+      "step": 20801
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3185353649116068e-05,
+      "loss": 1.6833,
+      "step": 20802
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3184785310599166e-05,
+      "loss": 1.8342,
+      "step": 20803
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3184216960633715e-05,
+      "loss": 1.8812,
+      "step": 20804
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.318364859922176e-05,
+      "loss": 1.6655,
+      "step": 20805
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3183080226365336e-05,
+      "loss": 1.9451,
+      "step": 20806
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3182511842066496e-05,
+      "loss": 1.9798,
+      "step": 20807
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3181943446327275e-05,
+      "loss": 1.9426,
+      "step": 20808
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3181375039149724e-05,
+      "loss": 1.7797,
+      "step": 20809
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3180806620535877e-05,
+      "loss": 1.9766,
+      "step": 20810
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3180238190487786e-05,
+      "loss": 1.8421,
+      "step": 20811
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3179669749007492e-05,
+      "loss": 1.8633,
+      "step": 20812
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3179101296097035e-05,
+      "loss": 1.8457,
+      "step": 20813
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.317853283175846e-05,
+      "loss": 2.0663,
+      "step": 20814
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3177964355993814e-05,
+      "loss": 1.7663,
+      "step": 20815
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3177395868805135e-05,
+      "loss": 1.6709,
+      "step": 20816
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3176827370194467e-05,
+      "loss": 1.9308,
+      "step": 20817
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.317625886016386e-05,
+      "loss": 1.6741,
+      "step": 20818
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3175690338715353e-05,
+      "loss": 1.8054,
+      "step": 20819
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.317512180585099e-05,
+      "loss": 2.0205,
+      "step": 20820
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.3174553261572816e-05,
+      "loss": 1.8684,
+      "step": 20821
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3173984705882871e-05,
+      "loss": 1.8483,
+      "step": 20822
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3173416138783205e-05,
+      "loss": 1.9389,
+      "step": 20823
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3172847560275854e-05,
+      "loss": 1.7355,
+      "step": 20824
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3172278970362868e-05,
+      "loss": 2.03,
+      "step": 20825
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3171710369046294e-05,
+      "loss": 1.8443,
+      "step": 20826
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3171141756328165e-05,
+      "loss": 2.0761,
+      "step": 20827
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3170573132210535e-05,
+      "loss": 2.2294,
+      "step": 20828
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3170004496695443e-05,
+      "loss": 1.6213,
+      "step": 20829
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3169435849784936e-05,
+      "loss": 1.635,
+      "step": 20830
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3168867191481054e-05,
+      "loss": 1.597,
+      "step": 20831
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3168298521785846e-05,
+      "loss": 1.7967,
+      "step": 20832
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.3167729840701356e-05,
+      "loss": 2.4579,
+      "step": 20833
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3167161148229622e-05,
+      "loss": 1.844,
+      "step": 20834
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3166592444372695e-05,
+      "loss": 1.6192,
+      "step": 20835
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3166023729132616e-05,
+      "loss": 2.1903,
+      "step": 20836
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3165455002511432e-05,
+      "loss": 1.6582,
+      "step": 20837
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3164886264511183e-05,
+      "loss": 1.8055,
+      "step": 20838
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.316431751513392e-05,
+      "loss": 1.764,
+      "step": 20839
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3163748754381682e-05,
+      "loss": 1.9602,
+      "step": 20840
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.3163179982256513e-05,
+      "loss": 2.1471,
+      "step": 20841
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3162611198760467e-05,
+      "loss": 2.0104,
+      "step": 20842
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.3162042403895574e-05,
+      "loss": 2.0544,
+      "step": 20843
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3161473597663892e-05,
+      "loss": 1.9262,
+      "step": 20844
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3160904780067456e-05,
+      "loss": 1.9151,
+      "step": 20845
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3160335951108318e-05,
+      "loss": 2.1126,
+      "step": 20846
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.315976711078852e-05,
+      "loss": 1.792,
+      "step": 20847
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3159198259110101e-05,
+      "loss": 1.6628,
+      "step": 20848
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3158629396075117e-05,
+      "loss": 1.6783,
+      "step": 20849
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.3158060521685604e-05,
+      "loss": 1.8215,
+      "step": 20850
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3157491635943612e-05,
+      "loss": 1.8269,
+      "step": 20851
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3156922738851186e-05,
+      "loss": 2.1223,
+      "step": 20852
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3156353830410366e-05,
+      "loss": 2.0229,
+      "step": 20853
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3155784910623202e-05,
+      "loss": 1.6959,
+      "step": 20854
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3155215979491736e-05,
+      "loss": 2.415,
+      "step": 20855
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.3154647037018018e-05,
+      "loss": 2.4575,
+      "step": 20856
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3154078083204087e-05,
+      "loss": 1.9777,
+      "step": 20857
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3153509118051993e-05,
+      "loss": 2.6753,
+      "step": 20858
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.315294014156378e-05,
+      "loss": 1.7426,
+      "step": 20859
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.315237115374149e-05,
+      "loss": 1.908,
+      "step": 20860
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3151802154587174e-05,
+      "loss": 2.058,
+      "step": 20861
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.315123314410287e-05,
+      "loss": 1.8102,
+      "step": 20862
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.3150664122290635e-05,
+      "loss": 1.8591,
+      "step": 20863
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3150095089152504e-05,
+      "loss": 2.0706,
+      "step": 20864
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3149526044690527e-05,
+      "loss": 2.0489,
+      "step": 20865
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3148956988906746e-05,
+      "loss": 1.9369,
+      "step": 20866
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3148387921803212e-05,
+      "loss": 1.5517,
+      "step": 20867
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3147818843381968e-05,
+      "loss": 1.9693,
+      "step": 20868
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.314724975364506e-05,
+      "loss": 1.7557,
+      "step": 20869
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3146680652594533e-05,
+      "loss": 1.7643,
+      "step": 20870
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.314611154023243e-05,
+      "loss": 1.9446,
+      "step": 20871
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3145542416560803e-05,
+      "loss": 1.9363,
+      "step": 20872
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3144973281581695e-05,
+      "loss": 2.0607,
+      "step": 20873
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.314440413529715e-05,
+      "loss": 1.9968,
+      "step": 20874
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3143834977709219e-05,
+      "loss": 1.725,
+      "step": 20875
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3143265808819943e-05,
+      "loss": 1.7936,
+      "step": 20876
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3142696628631371e-05,
+      "loss": 2.1773,
+      "step": 20877
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3142127437145545e-05,
+      "loss": 1.9434,
+      "step": 20878
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3141558234364516e-05,
+      "loss": 1.9043,
+      "step": 20879
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3140989020290327e-05,
+      "loss": 1.9228,
+      "step": 20880
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3140419794925026e-05,
+      "loss": 2.0451,
+      "step": 20881
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.313985055827066e-05,
+      "loss": 1.8778,
+      "step": 20882
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3139281310329271e-05,
+      "loss": 1.9461,
+      "step": 20883
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3138712051102908e-05,
+      "loss": 2.0692,
+      "step": 20884
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.313814278059362e-05,
+      "loss": 1.8459,
+      "step": 20885
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3137573498803451e-05,
+      "loss": 1.9228,
+      "step": 20886
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3137004205734444e-05,
+      "loss": 1.9388,
+      "step": 20887
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.313643490138865e-05,
+      "loss": 1.7381,
+      "step": 20888
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3135865585768116e-05,
+      "loss": 1.8548,
+      "step": 20889
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3135296258874885e-05,
+      "loss": 1.7308,
+      "step": 20890
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.313472692071101e-05,
+      "loss": 2.0308,
+      "step": 20891
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3134157571278525e-05,
+      "loss": 2.213,
+      "step": 20892
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3133588210579493e-05,
+      "loss": 1.7952,
+      "step": 20893
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3133018838615946e-05,
+      "loss": 1.6573,
+      "step": 20894
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.313244945538994e-05,
+      "loss": 1.9178,
+      "step": 20895
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.313188006090352e-05,
+      "loss": 1.5898,
+      "step": 20896
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.313131065515873e-05,
+      "loss": 2.1241,
+      "step": 20897
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3130741238157621e-05,
+      "loss": 1.6892,
+      "step": 20898
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3130171809902234e-05,
+      "loss": 2.0748,
+      "step": 20899
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3129602370394624e-05,
+      "loss": 1.9185,
+      "step": 20900
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3129032919636829e-05,
+      "loss": 1.4832,
+      "step": 20901
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3128463457630902e-05,
+      "loss": 1.5936,
+      "step": 20902
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.3127893984378892e-05,
+      "loss": 1.8387,
+      "step": 20903
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.312732449988284e-05,
+      "loss": 2.1441,
+      "step": 20904
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3126755004144796e-05,
+      "loss": 2.124,
+      "step": 20905
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3126185497166807e-05,
+      "loss": 1.6501,
+      "step": 20906
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3125615978950924e-05,
+      "loss": 2.0386,
+      "step": 20907
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3125046449499184e-05,
+      "loss": 1.902,
+      "step": 20908
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3124476908813645e-05,
+      "loss": 2.1532,
+      "step": 20909
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3123907356896352e-05,
+      "loss": 1.7716,
+      "step": 20910
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3123337793749348e-05,
+      "loss": 1.9892,
+      "step": 20911
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3122768219374686e-05,
+      "loss": 1.9726,
+      "step": 20912
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.312219863377441e-05,
+      "loss": 1.5602,
+      "step": 20913
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3121629036950568e-05,
+      "loss": 1.6634,
+      "step": 20914
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.3121059428905202e-05,
+      "loss": 1.7093,
+      "step": 20915
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3120489809640375e-05,
+      "loss": 1.5207,
+      "step": 20916
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3119920179158122e-05,
+      "loss": 1.9374,
+      "step": 20917
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.875,
+      "learning_rate": 1.3119350537460491e-05,
+      "loss": 2.1233,
+      "step": 20918
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3118780884549533e-05,
+      "loss": 1.7946,
+      "step": 20919
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.31182112204273e-05,
+      "loss": 2.008,
+      "step": 20920
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3117641545095829e-05,
+      "loss": 2.0711,
+      "step": 20921
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.3117071858557177e-05,
+      "loss": 1.9558,
+      "step": 20922
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3116502160813389e-05,
+      "loss": 1.965,
+      "step": 20923
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3115932451866514e-05,
+      "loss": 1.9067,
+      "step": 20924
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3115362731718594e-05,
+      "loss": 2.1646,
+      "step": 20925
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.3114793000371688e-05,
+      "loss": 1.887,
+      "step": 20926
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3114223257827832e-05,
+      "loss": 1.6828,
+      "step": 20927
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3113653504089083e-05,
+      "loss": 2.0287,
+      "step": 20928
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3113083739157486e-05,
+      "loss": 1.566,
+      "step": 20929
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3112513963035088e-05,
+      "loss": 2.0679,
+      "step": 20930
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.311194417572394e-05,
+      "loss": 1.9928,
+      "step": 20931
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.3111374377226084e-05,
+      "loss": 1.4958,
+      "step": 20932
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.311080456754358e-05,
+      "loss": 2.0099,
+      "step": 20933
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3110234746678466e-05,
+      "loss": 2.2487,
+      "step": 20934
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.310966491463279e-05,
+      "loss": 1.9927,
+      "step": 20935
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.310909507140861e-05,
+      "loss": 2.0047,
+      "step": 20936
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3108525217007966e-05,
+      "loss": 2.0172,
+      "step": 20937
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3107955351432908e-05,
+      "loss": 2.006,
+      "step": 20938
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3107385474685485e-05,
+      "loss": 1.8067,
+      "step": 20939
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3106815586767752e-05,
+      "loss": 1.694,
+      "step": 20940
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3106245687681745e-05,
+      "loss": 1.8012,
+      "step": 20941
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3105675777429521e-05,
+      "loss": 1.8638,
+      "step": 20942
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3105105856013127e-05,
+      "loss": 2.0932,
+      "step": 20943
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3104535923434613e-05,
+      "loss": 1.6981,
+      "step": 20944
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3103965979696028e-05,
+      "loss": 2.0411,
+      "step": 20945
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3103396024799415e-05,
+      "loss": 1.8374,
+      "step": 20946
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3102826058746831e-05,
+      "loss": 2.0972,
+      "step": 20947
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3102256081540319e-05,
+      "loss": 1.6796,
+      "step": 20948
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.310168609318193e-05,
+      "loss": 2.0455,
+      "step": 20949
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3101116093673715e-05,
+      "loss": 2.1218,
+      "step": 20950
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.310054608301772e-05,
+      "loss": 1.6388,
+      "step": 20951
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3099976061215995e-05,
+      "loss": 1.7351,
+      "step": 20952
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3099406028270589e-05,
+      "loss": 2.2183,
+      "step": 20953
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3098835984183553e-05,
+      "loss": 1.5515,
+      "step": 20954
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.3098265928956931e-05,
+      "loss": 2.0146,
+      "step": 20955
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3097695862592778e-05,
+      "loss": 1.8566,
+      "step": 20956
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3097125785093142e-05,
+      "loss": 1.7009,
+      "step": 20957
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3096555696460069e-05,
+      "loss": 1.6426,
+      "step": 20958
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3095985596695611e-05,
+      "loss": 2.0359,
+      "step": 20959
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3095415485801816e-05,
+      "loss": 1.9146,
+      "step": 20960
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.3094845363780739e-05,
+      "loss": 1.9115,
+      "step": 20961
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.309427523063442e-05,
+      "loss": 1.9845,
+      "step": 20962
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3093705086364916e-05,
+      "loss": 1.9478,
+      "step": 20963
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3093134930974271e-05,
+      "loss": 1.897,
+      "step": 20964
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.3092564764464539e-05,
+      "loss": 1.893,
+      "step": 20965
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3091994586837768e-05,
+      "loss": 1.875,
+      "step": 20966
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3091424398096009e-05,
+      "loss": 2.0659,
+      "step": 20967
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.309085419824131e-05,
+      "loss": 1.7099,
+      "step": 20968
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3090283987275719e-05,
+      "loss": 1.693,
+      "step": 20969
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.308971376520129e-05,
+      "loss": 1.6369,
+      "step": 20970
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.308914353202007e-05,
+      "loss": 2.0025,
+      "step": 20971
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.308857328773411e-05,
+      "loss": 1.9896,
+      "step": 20972
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.308800303234546e-05,
+      "loss": 1.8032,
+      "step": 20973
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3087432765856168e-05,
+      "loss": 2.0233,
+      "step": 20974
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3086862488268286e-05,
+      "loss": 2.0126,
+      "step": 20975
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3086292199583861e-05,
+      "loss": 2.0776,
+      "step": 20976
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3085721899804951e-05,
+      "loss": 1.9821,
+      "step": 20977
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3085151588933593e-05,
+      "loss": 1.8604,
+      "step": 20978
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3084581266971852e-05,
+      "loss": 2.0822,
+      "step": 20979
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3084010933921768e-05,
+      "loss": 1.9778,
+      "step": 20980
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3083440589785392e-05,
+      "loss": 1.9678,
+      "step": 20981
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.308287023456478e-05,
+      "loss": 1.9381,
+      "step": 20982
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3082299868261975e-05,
+      "loss": 1.8604,
+      "step": 20983
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3081729490879034e-05,
+      "loss": 1.6711,
+      "step": 20984
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3081159102418002e-05,
+      "loss": 1.5818,
+      "step": 20985
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3080588702880932e-05,
+      "loss": 2.1453,
+      "step": 20986
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3080018292269877e-05,
+      "loss": 2.3655,
+      "step": 20987
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3079447870586881e-05,
+      "loss": 1.7345,
+      "step": 20988
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3078877437833998e-05,
+      "loss": 1.7904,
+      "step": 20989
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3078306994013282e-05,
+      "loss": 1.8018,
+      "step": 20990
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3077736539126778e-05,
+      "loss": 1.9788,
+      "step": 20991
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3077166073176538e-05,
+      "loss": 2.1316,
+      "step": 20992
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3076595596164613e-05,
+      "loss": 1.935,
+      "step": 20993
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3076025108093063e-05,
+      "loss": 1.6351,
+      "step": 20994
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.307545460896392e-05,
+      "loss": 1.8569,
+      "step": 20995
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.307488409877925e-05,
+      "loss": 1.6818,
+      "step": 20996
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3074313577541097e-05,
+      "loss": 1.8417,
+      "step": 20997
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3073743045251515e-05,
+      "loss": 2.4269,
+      "step": 20998
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3073172501912552e-05,
+      "loss": 1.5547,
+      "step": 20999
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3072601947526264e-05,
+      "loss": 1.9535,
+      "step": 21000
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3072031382094697e-05,
+      "loss": 1.7597,
+      "step": 21001
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.30714608056199e-05,
+      "loss": 1.6702,
+      "step": 21002
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3070890218103931e-05,
+      "loss": 1.7345,
+      "step": 21003
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3070319619548839e-05,
+      "loss": 1.7288,
+      "step": 21004
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3069749009956674e-05,
+      "loss": 1.7484,
+      "step": 21005
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3069178389329487e-05,
+      "loss": 2.2327,
+      "step": 21006
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3068607757669328e-05,
+      "loss": 1.6778,
+      "step": 21007
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3068037114978252e-05,
+      "loss": 1.8085,
+      "step": 21008
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3067466461258302e-05,
+      "loss": 2.1257,
+      "step": 21009
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3066895796511543e-05,
+      "loss": 1.807,
+      "step": 21010
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.3066325120740016e-05,
+      "loss": 2.1532,
+      "step": 21011
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3065754433945777e-05,
+      "loss": 1.9928,
+      "step": 21012
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3065183736130873e-05,
+      "loss": 2.1171,
+      "step": 21013
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3064613027297359e-05,
+      "loss": 1.8075,
+      "step": 21014
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.3064042307447288e-05,
+      "loss": 1.79,
+      "step": 21015
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3063471576582707e-05,
+      "loss": 1.6111,
+      "step": 21016
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3062900834705672e-05,
+      "loss": 1.8505,
+      "step": 21017
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.306233008181823e-05,
+      "loss": 1.8574,
+      "step": 21018
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.306175931792244e-05,
+      "loss": 2.0225,
+      "step": 21019
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3061188543020347e-05,
+      "loss": 1.9156,
+      "step": 21020
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3060617757114005e-05,
+      "loss": 1.8088,
+      "step": 21021
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3060046960205466e-05,
+      "loss": 1.8303,
+      "step": 21022
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.305947615229678e-05,
+      "loss": 1.9388,
+      "step": 21023
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3058905333390006e-05,
+      "loss": 1.6253,
+      "step": 21024
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3058334503487185e-05,
+      "loss": 1.9383,
+      "step": 21025
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.3057763662590377e-05,
+      "loss": 1.9218,
+      "step": 21026
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3057192810701633e-05,
+      "loss": 2.1086,
+      "step": 21027
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3056621947823002e-05,
+      "loss": 1.9212,
+      "step": 21028
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3056051073956542e-05,
+      "loss": 1.8509,
+      "step": 21029
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3055480189104294e-05,
+      "loss": 1.9256,
+      "step": 21030
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3054909293268323e-05,
+      "loss": 2.0114,
+      "step": 21031
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3054338386450673e-05,
+      "loss": 1.7218,
+      "step": 21032
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3053767468653399e-05,
+      "loss": 1.9004,
+      "step": 21033
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3053196539878554e-05,
+      "loss": 1.4663,
+      "step": 21034
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3052625600128192e-05,
+      "loss": 2.0786,
+      "step": 21035
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.305205464940436e-05,
+      "loss": 2.1597,
+      "step": 21036
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3051483687709112e-05,
+      "loss": 1.7569,
+      "step": 21037
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.3050912715044506e-05,
+      "loss": 1.589,
+      "step": 21038
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.3050341731412586e-05,
+      "loss": 1.8139,
+      "step": 21039
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.304977073681541e-05,
+      "loss": 2.038,
+      "step": 21040
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3049199731255028e-05,
+      "loss": 1.8174,
+      "step": 21041
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3048628714733498e-05,
+      "loss": 2.3192,
+      "step": 21042
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3048057687252866e-05,
+      "loss": 1.8998,
+      "step": 21043
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.3047486648815191e-05,
+      "loss": 1.9347,
+      "step": 21044
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3046915599422518e-05,
+      "loss": 2.0562,
+      "step": 21045
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3046344539076905e-05,
+      "loss": 1.8884,
+      "step": 21046
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3045773467780407e-05,
+      "loss": 2.249,
+      "step": 21047
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.304520238553507e-05,
+      "loss": 1.8914,
+      "step": 21048
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3044631292342951e-05,
+      "loss": 2.0526,
+      "step": 21049
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3044060188206103e-05,
+      "loss": 1.7142,
+      "step": 21050
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3043489073126578e-05,
+      "loss": 1.8381,
+      "step": 21051
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3042917947106432e-05,
+      "loss": 2.2464,
+      "step": 21052
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.304234681014771e-05,
+      "loss": 1.7256,
+      "step": 21053
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3041775662252479e-05,
+      "loss": 2.1676,
+      "step": 21054
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3041204503422778e-05,
+      "loss": 1.7762,
+      "step": 21055
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3040633333660666e-05,
+      "loss": 2.203,
+      "step": 21056
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3040062152968197e-05,
+      "loss": 1.8268,
+      "step": 21057
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3039490961347425e-05,
+      "loss": 1.9747,
+      "step": 21058
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3038919758800402e-05,
+      "loss": 1.7434,
+      "step": 21059
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3038348545329178e-05,
+      "loss": 1.8513,
+      "step": 21060
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3037777320935815e-05,
+      "loss": 1.6242,
+      "step": 21061
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3037206085622355e-05,
+      "loss": 1.652,
+      "step": 21062
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.303663483939086e-05,
+      "loss": 1.7954,
+      "step": 21063
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3036063582243381e-05,
+      "loss": 1.807,
+      "step": 21064
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.3035492314181973e-05,
+      "loss": 1.8642,
+      "step": 21065
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.3034921035208688e-05,
+      "loss": 1.6969,
+      "step": 21066
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.3034349745325577e-05,
+      "loss": 2.4512,
+      "step": 21067
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3033778444534697e-05,
+      "loss": 2.0894,
+      "step": 21068
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.30332071328381e-05,
+      "loss": 2.6412,
+      "step": 21069
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.3032635810237841e-05,
+      "loss": 1.8558,
+      "step": 21070
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.3032064476735977e-05,
+      "loss": 2.4336,
+      "step": 21071
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3031493132334554e-05,
+      "loss": 2.1877,
+      "step": 21072
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3030921777035632e-05,
+      "loss": 1.6929,
+      "step": 21073
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3030350410841264e-05,
+      "loss": 1.7157,
+      "step": 21074
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.30297790337535e-05,
+      "loss": 2.0343,
+      "step": 21075
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.3029207645774398e-05,
+      "loss": 1.769,
+      "step": 21076
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.302863624690601e-05,
+      "loss": 1.5872,
+      "step": 21077
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3028064837150397e-05,
+      "loss": 1.9291,
+      "step": 21078
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.30274934165096e-05,
+      "loss": 1.6654,
+      "step": 21079
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3026921984985683e-05,
+      "loss": 1.772,
+      "step": 21080
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3026350542580698e-05,
+      "loss": 1.9077,
+      "step": 21081
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3025779089296698e-05,
+      "loss": 2.2215,
+      "step": 21082
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3025207625135734e-05,
+      "loss": 1.577,
+      "step": 21083
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.302463615009987e-05,
+      "loss": 2.0456,
+      "step": 21084
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3024064664191151e-05,
+      "loss": 2.052,
+      "step": 21085
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3023493167411635e-05,
+      "loss": 1.896,
+      "step": 21086
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.3022921659763376e-05,
+      "loss": 2.1757,
+      "step": 21087
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.3022350141248428e-05,
+      "loss": 2.0818,
+      "step": 21088
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.125,
+      "learning_rate": 1.3021778611868847e-05,
+      "loss": 2.1843,
+      "step": 21089
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3021207071626687e-05,
+      "loss": 1.9459,
+      "step": 21090
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3020635520524004e-05,
+      "loss": 2.2671,
+      "step": 21091
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3020063958562847e-05,
+      "loss": 1.6676,
+      "step": 21092
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.3019492385745273e-05,
+      "loss": 1.7572,
+      "step": 21093
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.3018920802073343e-05,
+      "loss": 1.7156,
+      "step": 21094
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.30183492075491e-05,
+      "loss": 1.6291,
+      "step": 21095
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3017777602174612e-05,
+      "loss": 2.4995,
+      "step": 21096
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3017205985951926e-05,
+      "loss": 2.041,
+      "step": 21097
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3016634358883096e-05,
+      "loss": 2.0347,
+      "step": 21098
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3016062720970179e-05,
+      "loss": 2.0493,
+      "step": 21099
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.301549107221523e-05,
+      "loss": 2.1098,
+      "step": 21100
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.3014919412620306e-05,
+      "loss": 1.9002,
+      "step": 21101
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.3014347742187457e-05,
+      "loss": 1.8451,
+      "step": 21102
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.3013776060918742e-05,
+      "loss": 2.2977,
+      "step": 21103
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3013204368816213e-05,
+      "loss": 2.2657,
+      "step": 21104
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.301263266588193e-05,
+      "loss": 1.715,
+      "step": 21105
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.3012060952117943e-05,
+      "loss": 1.7425,
+      "step": 21106
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3011489227526308e-05,
+      "loss": 1.7992,
+      "step": 21107
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.3010917492109086e-05,
+      "loss": 1.6287,
+      "step": 21108
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3010345745868322e-05,
+      "loss": 1.8132,
+      "step": 21109
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.3009773988806081e-05,
+      "loss": 1.8588,
+      "step": 21110
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.300920222092441e-05,
+      "loss": 1.6888,
+      "step": 21111
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.3008630442225375e-05,
+      "loss": 2.0039,
+      "step": 21112
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.300805865271102e-05,
+      "loss": 1.9667,
+      "step": 21113
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3007486852383406e-05,
+      "loss": 1.7204,
+      "step": 21114
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.3006915041244592e-05,
+      "loss": 2.1744,
+      "step": 21115
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3006343219296624e-05,
+      "loss": 1.8627,
+      "step": 21116
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.3005771386541569e-05,
+      "loss": 1.8586,
+      "step": 21117
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3005199542981473e-05,
+      "loss": 1.6906,
+      "step": 21118
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.3004627688618394e-05,
+      "loss": 1.8797,
+      "step": 21119
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.375,
+      "learning_rate": 1.3004055823454391e-05,
+      "loss": 2.1412,
+      "step": 21120
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.300348394749152e-05,
+      "loss": 1.7814,
+      "step": 21121
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.3002912060731832e-05,
+      "loss": 1.625,
+      "step": 21122
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3002340163177382e-05,
+      "loss": 2.2066,
+      "step": 21123
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3001768254830234e-05,
+      "loss": 1.9573,
+      "step": 21124
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.300119633569244e-05,
+      "loss": 2.0323,
+      "step": 21125
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.300062440576605e-05,
+      "loss": 1.9583,
+      "step": 21126
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.3000052465053128e-05,
+      "loss": 1.5389,
+      "step": 21127
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2999480513555729e-05,
+      "loss": 1.9067,
+      "step": 21128
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2998908551275904e-05,
+      "loss": 1.957,
+      "step": 21129
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2998336578215709e-05,
+      "loss": 2.0876,
+      "step": 21130
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.299776459437721e-05,
+      "loss": 1.7121,
+      "step": 21131
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2997192599762452e-05,
+      "loss": 1.8828,
+      "step": 21132
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2996620594373495e-05,
+      "loss": 1.9314,
+      "step": 21133
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2996048578212396e-05,
+      "loss": 1.7471,
+      "step": 21134
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2995476551281211e-05,
+      "loss": 1.7138,
+      "step": 21135
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2994904513581998e-05,
+      "loss": 1.8684,
+      "step": 21136
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.299433246511681e-05,
+      "loss": 1.9707,
+      "step": 21137
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2993760405887707e-05,
+      "loss": 1.774,
+      "step": 21138
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2993188335896741e-05,
+      "loss": 1.8505,
+      "step": 21139
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2992616255145972e-05,
+      "loss": 1.6011,
+      "step": 21140
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2992044163637454e-05,
+      "loss": 2.605,
+      "step": 21141
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2991472061373245e-05,
+      "loss": 1.9684,
+      "step": 21142
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2990899948355403e-05,
+      "loss": 1.7841,
+      "step": 21143
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.299032782458598e-05,
+      "loss": 2.038,
+      "step": 21144
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2989755690067042e-05,
+      "loss": 1.8391,
+      "step": 21145
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.298918354480063e-05,
+      "loss": 1.8456,
+      "step": 21146
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2988611388788817e-05,
+      "loss": 2.1926,
+      "step": 21147
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2988039222033653e-05,
+      "loss": 1.7075,
+      "step": 21148
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.298746704453719e-05,
+      "loss": 1.7469,
+      "step": 21149
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2986894856301493e-05,
+      "loss": 2.059,
+      "step": 21150
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2986322657328614e-05,
+      "loss": 2.2131,
+      "step": 21151
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2985750447620612e-05,
+      "loss": 1.8576,
+      "step": 21152
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2985178227179543e-05,
+      "loss": 2.161,
+      "step": 21153
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2984605996007462e-05,
+      "loss": 1.8965,
+      "step": 21154
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2984033754106433e-05,
+      "loss": 1.8582,
+      "step": 21155
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2983461501478504e-05,
+      "loss": 1.7283,
+      "step": 21156
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.298288923812574e-05,
+      "loss": 1.9522,
+      "step": 21157
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2982316964050193e-05,
+      "loss": 1.813,
+      "step": 21158
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.2981744679253923e-05,
+      "loss": 1.6216,
+      "step": 21159
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.298117238373898e-05,
+      "loss": 2.0348,
+      "step": 21160
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2980600077507432e-05,
+      "loss": 1.405,
+      "step": 21161
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2980027760561336e-05,
+      "loss": 1.7102,
+      "step": 21162
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2979455432902739e-05,
+      "loss": 1.7732,
+      "step": 21163
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2978883094533706e-05,
+      "loss": 1.8929,
+      "step": 21164
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2978310745456293e-05,
+      "loss": 1.8461,
+      "step": 21165
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2977738385672558e-05,
+      "loss": 1.8227,
+      "step": 21166
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2977166015184557e-05,
+      "loss": 1.7311,
+      "step": 21167
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2976593633994347e-05,
+      "loss": 1.9771,
+      "step": 21168
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2976021242103988e-05,
+      "loss": 1.7862,
+      "step": 21169
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2975448839515536e-05,
+      "loss": 1.9663,
+      "step": 21170
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.297487642623105e-05,
+      "loss": 1.9393,
+      "step": 21171
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2974304002252584e-05,
+      "loss": 1.6266,
+      "step": 21172
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2973731567582201e-05,
+      "loss": 2.0027,
+      "step": 21173
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2973159122221956e-05,
+      "loss": 1.9492,
+      "step": 21174
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.297258666617391e-05,
+      "loss": 2.0616,
+      "step": 21175
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2972014199440115e-05,
+      "loss": 1.8721,
+      "step": 21176
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2971441722022629e-05,
+      "loss": 1.8606,
+      "step": 21177
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2970869233923517e-05,
+      "loss": 1.8695,
+      "step": 21178
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2970296735144829e-05,
+      "loss": 2.0006,
+      "step": 21179
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.375,
+      "learning_rate": 1.296972422568863e-05,
+      "loss": 2.0451,
+      "step": 21180
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2969151705556975e-05,
+      "loss": 1.8115,
+      "step": 21181
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.296857917475192e-05,
+      "loss": 1.634,
+      "step": 21182
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2968006633275525e-05,
+      "loss": 1.9189,
+      "step": 21183
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2967434081129846e-05,
+      "loss": 1.8394,
+      "step": 21184
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.296686151831695e-05,
+      "loss": 1.7507,
+      "step": 21185
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2966288944838881e-05,
+      "loss": 1.5693,
+      "step": 21186
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2965716360697709e-05,
+      "loss": 1.7209,
+      "step": 21187
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2965143765895487e-05,
+      "loss": 1.9768,
+      "step": 21188
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2964571160434275e-05,
+      "loss": 1.7416,
+      "step": 21189
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.296399854431613e-05,
+      "loss": 1.8279,
+      "step": 21190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.296342591754311e-05,
+      "loss": 1.7829,
+      "step": 21191
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2962853280117278e-05,
+      "loss": 1.9485,
+      "step": 21192
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2962280632040684e-05,
+      "loss": 1.9579,
+      "step": 21193
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2961707973315396e-05,
+      "loss": 1.8187,
+      "step": 21194
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2961135303943468e-05,
+      "loss": 1.6199,
+      "step": 21195
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2960562623926958e-05,
+      "loss": 1.606,
+      "step": 21196
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2959989933267924e-05,
+      "loss": 1.7548,
+      "step": 21197
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.295941723196843e-05,
+      "loss": 2.2065,
+      "step": 21198
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2958844520030529e-05,
+      "loss": 1.9199,
+      "step": 21199
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2958271797456278e-05,
+      "loss": 2.0053,
+      "step": 21200
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2957699064247744e-05,
+      "loss": 1.7771,
+      "step": 21201
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2957126320406981e-05,
+      "loss": 1.6508,
+      "step": 21202
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2956553565936045e-05,
+      "loss": 1.8218,
+      "step": 21203
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2955980800837003e-05,
+      "loss": 2.1137,
+      "step": 21204
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2955408025111906e-05,
+      "loss": 1.6811,
+      "step": 21205
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2954835238762817e-05,
+      "loss": 1.7584,
+      "step": 21206
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2954262441791792e-05,
+      "loss": 2.1509,
+      "step": 21207
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2953689634200894e-05,
+      "loss": 2.0207,
+      "step": 21208
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2953116815992182e-05,
+      "loss": 1.9297,
+      "step": 21209
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2952543987167712e-05,
+      "loss": 1.9884,
+      "step": 21210
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.2951971147729542e-05,
+      "loss": 2.1651,
+      "step": 21211
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2951398297679738e-05,
+      "loss": 2.1997,
+      "step": 21212
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2950825437020354e-05,
+      "loss": 1.4334,
+      "step": 21213
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2950252565753447e-05,
+      "loss": 2.2148,
+      "step": 21214
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2949679683881084e-05,
+      "loss": 1.8212,
+      "step": 21215
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2949106791405316e-05,
+      "loss": 1.823,
+      "step": 21216
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2948533888328212e-05,
+      "loss": 1.9922,
+      "step": 21217
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2947960974651823e-05,
+      "loss": 1.8659,
+      "step": 21218
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.294738805037821e-05,
+      "loss": 1.8751,
+      "step": 21219
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.125,
+      "learning_rate": 1.2946815115509436e-05,
+      "loss": 1.9872,
+      "step": 21220
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2946242170047558e-05,
+      "loss": 1.8124,
+      "step": 21221
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2945669213994638e-05,
+      "loss": 1.8597,
+      "step": 21222
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2945096247352731e-05,
+      "loss": 2.0404,
+      "step": 21223
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2944523270123902e-05,
+      "loss": 2.1259,
+      "step": 21224
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2943950282310205e-05,
+      "loss": 1.6651,
+      "step": 21225
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2943377283913705e-05,
+      "loss": 1.4764,
+      "step": 21226
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2942804274936461e-05,
+      "loss": 2.0688,
+      "step": 21227
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2942231255380526e-05,
+      "loss": 2.0723,
+      "step": 21228
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2941658225247972e-05,
+      "loss": 1.9401,
+      "step": 21229
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.2941085184540848e-05,
+      "loss": 2.3736,
+      "step": 21230
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2940512133261222e-05,
+      "loss": 2.2064,
+      "step": 21231
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.293993907141115e-05,
+      "loss": 1.9234,
+      "step": 21232
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.293936599899269e-05,
+      "loss": 1.5809,
+      "step": 21233
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2938792916007904e-05,
+      "loss": 2.152,
+      "step": 21234
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2938219822458856e-05,
+      "loss": 1.8738,
+      "step": 21235
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2937646718347598e-05,
+      "loss": 1.7142,
+      "step": 21236
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2937073603676195e-05,
+      "loss": 1.9906,
+      "step": 21237
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.293650047844671e-05,
+      "loss": 1.9026,
+      "step": 21238
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2935927342661204e-05,
+      "loss": 1.9513,
+      "step": 21239
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2935354196321726e-05,
+      "loss": 1.8484,
+      "step": 21240
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2934781039430346e-05,
+      "loss": 1.9356,
+      "step": 21241
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2934207871989121e-05,
+      "loss": 1.9919,
+      "step": 21242
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2933634694000117e-05,
+      "loss": 1.9676,
+      "step": 21243
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2933061505465385e-05,
+      "loss": 2.0397,
+      "step": 21244
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2932488306386995e-05,
+      "loss": 2.053,
+      "step": 21245
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.2931915096767002e-05,
+      "loss": 1.6695,
+      "step": 21246
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2931341876607463e-05,
+      "loss": 1.6704,
+      "step": 21247
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.293076864591045e-05,
+      "loss": 1.7396,
+      "step": 21248
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2930195404678014e-05,
+      "loss": 1.908,
+      "step": 21249
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2929622152912218e-05,
+      "loss": 1.8258,
+      "step": 21250
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2929048890615124e-05,
+      "loss": 1.7346,
+      "step": 21251
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2928475617788791e-05,
+      "loss": 1.6849,
+      "step": 21252
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2927902334435283e-05,
+      "loss": 1.5088,
+      "step": 21253
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2927329040556652e-05,
+      "loss": 1.7146,
+      "step": 21254
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2926755736154972e-05,
+      "loss": 1.8314,
+      "step": 21255
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2926182421232296e-05,
+      "loss": 1.6743,
+      "step": 21256
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2925609095790685e-05,
+      "loss": 1.8346,
+      "step": 21257
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2925035759832203e-05,
+      "loss": 1.8438,
+      "step": 21258
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2924462413358908e-05,
+      "loss": 1.9824,
+      "step": 21259
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2923889056372863e-05,
+      "loss": 2.2082,
+      "step": 21260
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2923315688876126e-05,
+      "loss": 1.6905,
+      "step": 21261
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2922742310870762e-05,
+      "loss": 1.4564,
+      "step": 21262
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2922168922358829e-05,
+      "loss": 1.8631,
+      "step": 21263
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2921595523342391e-05,
+      "loss": 2.1284,
+      "step": 21264
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2921022113823509e-05,
+      "loss": 1.4826,
+      "step": 21265
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2920448693804244e-05,
+      "loss": 1.7651,
+      "step": 21266
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2919875263286654e-05,
+      "loss": 1.5336,
+      "step": 21267
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2919301822272803e-05,
+      "loss": 2.0515,
+      "step": 21268
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2918728370764756e-05,
+      "loss": 1.4633,
+      "step": 21269
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2918154908764566e-05,
+      "loss": 1.8969,
+      "step": 21270
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.2917581436274302e-05,
+      "loss": 1.908,
+      "step": 21271
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2917007953296023e-05,
+      "loss": 1.9503,
+      "step": 21272
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.291643445983179e-05,
+      "loss": 1.9706,
+      "step": 21273
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2915860955883664e-05,
+      "loss": 1.9209,
+      "step": 21274
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2915287441453708e-05,
+      "loss": 1.6112,
+      "step": 21275
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2914713916543983e-05,
+      "loss": 1.8149,
+      "step": 21276
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.291414038115655e-05,
+      "loss": 1.6004,
+      "step": 21277
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2913566835293472e-05,
+      "loss": 2.0229,
+      "step": 21278
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2912993278956812e-05,
+      "loss": 1.9815,
+      "step": 21279
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2912419712148628e-05,
+      "loss": 2.0703,
+      "step": 21280
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2911846134870984e-05,
+      "loss": 2.0975,
+      "step": 21281
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2911272547125944e-05,
+      "loss": 1.8757,
+      "step": 21282
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2910698948915565e-05,
+      "loss": 1.8183,
+      "step": 21283
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.291012534024191e-05,
+      "loss": 1.9216,
+      "step": 21284
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2909551721107048e-05,
+      "loss": 1.783,
+      "step": 21285
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.290897809151303e-05,
+      "loss": 1.7409,
+      "step": 21286
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.290840445146193e-05,
+      "loss": 1.9274,
+      "step": 21287
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2907830800955799e-05,
+      "loss": 2.3337,
+      "step": 21288
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2907257139996704e-05,
+      "loss": 1.7407,
+      "step": 21289
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2906683468586709e-05,
+      "loss": 1.8458,
+      "step": 21290
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2906109786727874e-05,
+      "loss": 1.9225,
+      "step": 21291
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2905536094422261e-05,
+      "loss": 1.9164,
+      "step": 21292
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2904962391671933e-05,
+      "loss": 2.0999,
+      "step": 21293
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2904388678478954e-05,
+      "loss": 2.1662,
+      "step": 21294
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2903814954845381e-05,
+      "loss": 2.098,
+      "step": 21295
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2903241220773284e-05,
+      "loss": 1.7437,
+      "step": 21296
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.290266747626472e-05,
+      "loss": 3.0272,
+      "step": 21297
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2902093721321749e-05,
+      "loss": 1.9923,
+      "step": 21298
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.2901519955946445e-05,
+      "loss": 0.1132,
+      "step": 21299
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2900946180140856e-05,
+      "loss": 1.5353,
+      "step": 21300
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2900372393907054e-05,
+      "loss": 1.9594,
+      "step": 21301
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.28997985972471e-05,
+      "loss": 1.6956,
+      "step": 21302
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2899224790163056e-05,
+      "loss": 1.7862,
+      "step": 21303
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2898650972656984e-05,
+      "loss": 1.9076,
+      "step": 21304
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2898077144730946e-05,
+      "loss": 1.808,
+      "step": 21305
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2897503306387007e-05,
+      "loss": 1.9281,
+      "step": 21306
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2896929457627227e-05,
+      "loss": 1.6341,
+      "step": 21307
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2896355598453674e-05,
+      "loss": 1.7669,
+      "step": 21308
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2895781728868408e-05,
+      "loss": 1.9739,
+      "step": 21309
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2895207848873488e-05,
+      "loss": 1.6841,
+      "step": 21310
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2894633958470984e-05,
+      "loss": 1.6574,
+      "step": 21311
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.289406005766295e-05,
+      "loss": 1.8848,
+      "step": 21312
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2893486146451459e-05,
+      "loss": 2.0426,
+      "step": 21313
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2892912224838568e-05,
+      "loss": 1.7418,
+      "step": 21314
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2892338292826343e-05,
+      "loss": 1.6643,
+      "step": 21315
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2891764350416848e-05,
+      "loss": 1.7197,
+      "step": 21316
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2891190397612138e-05,
+      "loss": 1.8347,
+      "step": 21317
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2890616434414286e-05,
+      "loss": 1.8001,
+      "step": 21318
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2890042460825351e-05,
+      "loss": 1.8437,
+      "step": 21319
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.28894684768474e-05,
+      "loss": 1.791,
+      "step": 21320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2888894482482486e-05,
+      "loss": 1.5951,
+      "step": 21321
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2888320477732686e-05,
+      "loss": 1.8723,
+      "step": 21322
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2887746462600057e-05,
+      "loss": 1.9118,
+      "step": 21323
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2887172437086657e-05,
+      "loss": 2.028,
+      "step": 21324
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.288659840119456e-05,
+      "loss": 2.1133,
+      "step": 21325
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2886024354925821e-05,
+      "loss": 1.8733,
+      "step": 21326
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2885450298282509e-05,
+      "loss": 1.8208,
+      "step": 21327
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2884876231266683e-05,
+      "loss": 1.9421,
+      "step": 21328
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.288430215388041e-05,
+      "loss": 2.1036,
+      "step": 21329
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2883728066125758e-05,
+      "loss": 1.6426,
+      "step": 21330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2883153968004779e-05,
+      "loss": 2.1769,
+      "step": 21331
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2882579859519546e-05,
+      "loss": 1.6089,
+      "step": 21332
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.288200574067212e-05,
+      "loss": 1.8586,
+      "step": 21333
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2881431611464565e-05,
+      "loss": 1.8592,
+      "step": 21334
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2880857471898946e-05,
+      "loss": 1.9096,
+      "step": 21335
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2880283321977323e-05,
+      "loss": 1.9467,
+      "step": 21336
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.2879709161701765e-05,
+      "loss": 1.7515,
+      "step": 21337
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2879134991074332e-05,
+      "loss": 1.6724,
+      "step": 21338
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.287856081009709e-05,
+      "loss": 1.9684,
+      "step": 21339
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2877986618772102e-05,
+      "loss": 1.8051,
+      "step": 21340
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2877412417101433e-05,
+      "loss": 1.8817,
+      "step": 21341
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.625,
+      "learning_rate": 1.287683820508715e-05,
+      "loss": 2.5137,
+      "step": 21342
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.287626398273131e-05,
+      "loss": 2.2242,
+      "step": 21343
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2875689750035982e-05,
+      "loss": 2.8782,
+      "step": 21344
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.287511550700323e-05,
+      "loss": 1.6594,
+      "step": 21345
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.287454125363512e-05,
+      "loss": 1.6946,
+      "step": 21346
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.287396698993371e-05,
+      "loss": 1.9839,
+      "step": 21347
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.287339271590107e-05,
+      "loss": 1.8757,
+      "step": 21348
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2872818431539264e-05,
+      "loss": 1.8436,
+      "step": 21349
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2872244136850354e-05,
+      "loss": 1.7222,
+      "step": 21350
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2871669831836406e-05,
+      "loss": 1.9531,
+      "step": 21351
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.287109551649948e-05,
+      "loss": 1.909,
+      "step": 21352
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2870521190841652e-05,
+      "loss": 2.1679,
+      "step": 21353
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2869946854864974e-05,
+      "loss": 1.8361,
+      "step": 21354
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2869372508571517e-05,
+      "loss": 1.9161,
+      "step": 21355
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2868798151963344e-05,
+      "loss": 1.7787,
+      "step": 21356
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2868223785042522e-05,
+      "loss": 1.7545,
+      "step": 21357
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2867649407811113e-05,
+      "loss": 1.5357,
+      "step": 21358
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.286707502027118e-05,
+      "loss": 1.9099,
+      "step": 21359
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2866500622424795e-05,
+      "loss": 1.8854,
+      "step": 21360
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2865926214274014e-05,
+      "loss": 2.3662,
+      "step": 21361
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2865351795820908e-05,
+      "loss": 2.071,
+      "step": 21362
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.286477736706754e-05,
+      "loss": 2.3519,
+      "step": 21363
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2864202928015973e-05,
+      "loss": 1.6866,
+      "step": 21364
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2863628478668278e-05,
+      "loss": 1.7914,
+      "step": 21365
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2863054019026513e-05,
+      "loss": 2.0297,
+      "step": 21366
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2862479549092745e-05,
+      "loss": 1.7821,
+      "step": 21367
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.286190506886904e-05,
+      "loss": 1.6718,
+      "step": 21368
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2861330578357465e-05,
+      "loss": 1.8412,
+      "step": 21369
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2860756077560081e-05,
+      "loss": 1.8597,
+      "step": 21370
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2860181566478958e-05,
+      "loss": 1.7672,
+      "step": 21371
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2859607045116158e-05,
+      "loss": 2.0989,
+      "step": 21372
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2859032513473745e-05,
+      "loss": 1.6727,
+      "step": 21373
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2858457971553788e-05,
+      "loss": 2.0272,
+      "step": 21374
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.285788341935835e-05,
+      "loss": 1.8294,
+      "step": 21375
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.2857308856889499e-05,
+      "loss": 2.1123,
+      "step": 21376
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2856734284149294e-05,
+      "loss": 1.9857,
+      "step": 21377
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.285615970113981e-05,
+      "loss": 1.834,
+      "step": 21378
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2855585107863105e-05,
+      "loss": 1.8306,
+      "step": 21379
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2855010504321245e-05,
+      "loss": 2.2376,
+      "step": 21380
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2854435890516301e-05,
+      "loss": 1.7961,
+      "step": 21381
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2853861266450332e-05,
+      "loss": 1.8557,
+      "step": 21382
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2853286632125411e-05,
+      "loss": 1.8287,
+      "step": 21383
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2852711987543594e-05,
+      "loss": 1.8761,
+      "step": 21384
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2852137332706954e-05,
+      "loss": 1.8601,
+      "step": 21385
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2851562667617554e-05,
+      "loss": 1.8011,
+      "step": 21386
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2850987992277463e-05,
+      "loss": 2.3614,
+      "step": 21387
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2850413306688743e-05,
+      "loss": 2.0356,
+      "step": 21388
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.284983861085346e-05,
+      "loss": 2.0024,
+      "step": 21389
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2849263904773684e-05,
+      "loss": 1.9889,
+      "step": 21390
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2848689188451476e-05,
+      "loss": 1.6986,
+      "step": 21391
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2848114461888902e-05,
+      "loss": 1.9054,
+      "step": 21392
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2847539725088037e-05,
+      "loss": 1.6657,
+      "step": 21393
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2846964978050934e-05,
+      "loss": 1.7413,
+      "step": 21394
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2846390220779667e-05,
+      "loss": 1.9938,
+      "step": 21395
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2845815453276298e-05,
+      "loss": 1.7551,
+      "step": 21396
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.375,
+      "learning_rate": 1.28452406755429e-05,
+      "loss": 2.0332,
+      "step": 21397
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2844665887581531e-05,
+      "loss": 1.6987,
+      "step": 21398
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2844091089394261e-05,
+      "loss": 1.7905,
+      "step": 21399
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2843516280983162e-05,
+      "loss": 2.0587,
+      "step": 21400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2842941462350287e-05,
+      "loss": 1.7546,
+      "step": 21401
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2842366633497714e-05,
+      "loss": 1.926,
+      "step": 21402
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2841791794427505e-05,
+      "loss": 1.9113,
+      "step": 21403
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2841216945141723e-05,
+      "loss": 1.7669,
+      "step": 21404
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.2840642085642441e-05,
+      "loss": 1.8293,
+      "step": 21405
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.2840067215931721e-05,
+      "loss": 2.545,
+      "step": 21406
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2839492336011634e-05,
+      "loss": 1.7131,
+      "step": 21407
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.283891744588424e-05,
+      "loss": 1.7201,
+      "step": 21408
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2838342545551612e-05,
+      "loss": 1.7557,
+      "step": 21409
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2837767635015812e-05,
+      "loss": 2.0885,
+      "step": 21410
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.375,
+      "learning_rate": 1.283719271427891e-05,
+      "loss": 2.0423,
+      "step": 21411
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2836617783342968e-05,
+      "loss": 2.1222,
+      "step": 21412
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2836042842210057e-05,
+      "loss": 1.6374,
+      "step": 21413
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.2835467890882244e-05,
+      "loss": 2.1917,
+      "step": 21414
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2834892929361595e-05,
+      "loss": 1.7728,
+      "step": 21415
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2834317957650175e-05,
+      "loss": 2.0496,
+      "step": 21416
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2833742975750055e-05,
+      "loss": 2.2986,
+      "step": 21417
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2833167983663294e-05,
+      "loss": 1.6891,
+      "step": 21418
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.283259298139197e-05,
+      "loss": 2.1062,
+      "step": 21419
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2832017968938139e-05,
+      "loss": 1.5357,
+      "step": 21420
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.2831442946303878e-05,
+      "loss": 2.0085,
+      "step": 21421
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2830867913491245e-05,
+      "loss": 1.8332,
+      "step": 21422
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2830292870502312e-05,
+      "loss": 1.7682,
+      "step": 21423
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2829717817339147e-05,
+      "loss": 1.666,
+      "step": 21424
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2829142754003815e-05,
+      "loss": 1.8254,
+      "step": 21425
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2828567680498383e-05,
+      "loss": 1.7446,
+      "step": 21426
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2827992596824922e-05,
+      "loss": 1.9017,
+      "step": 21427
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2827417502985493e-05,
+      "loss": 1.841,
+      "step": 21428
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2826842398982166e-05,
+      "loss": 1.8397,
+      "step": 21429
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2826267284817015e-05,
+      "loss": 1.7524,
+      "step": 21430
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2825692160492096e-05,
+      "loss": 1.7878,
+      "step": 21431
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2825117026009483e-05,
+      "loss": 1.8511,
+      "step": 21432
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2824541881371244e-05,
+      "loss": 2.0187,
+      "step": 21433
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2823966726579444e-05,
+      "loss": 1.8732,
+      "step": 21434
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2823391561636153e-05,
+      "loss": 1.8671,
+      "step": 21435
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2822816386543434e-05,
+      "loss": 1.9467,
+      "step": 21436
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2822241201303361e-05,
+      "loss": 2.0894,
+      "step": 21437
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2821666005917994e-05,
+      "loss": 1.9744,
+      "step": 21438
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2821090800389409e-05,
+      "loss": 1.9346,
+      "step": 21439
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2820515584719669e-05,
+      "loss": 2.1529,
+      "step": 21440
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.281994035891084e-05,
+      "loss": 1.8698,
+      "step": 21441
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2819365122964995e-05,
+      "loss": 2.0797,
+      "step": 21442
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2818789876884197e-05,
+      "loss": 1.8558,
+      "step": 21443
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.281821462067052e-05,
+      "loss": 1.8697,
+      "step": 21444
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2817639354326024e-05,
+      "loss": 2.0517,
+      "step": 21445
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2817064077852781e-05,
+      "loss": 1.8071,
+      "step": 21446
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2816488791252862e-05,
+      "loss": 2.185,
+      "step": 21447
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.2815913494528329e-05,
+      "loss": 1.9595,
+      "step": 21448
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2815338187681254e-05,
+      "loss": 1.7252,
+      "step": 21449
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2814762870713704e-05,
+      "loss": 1.7647,
+      "step": 21450
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2814187543627747e-05,
+      "loss": 1.8224,
+      "step": 21451
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.281361220642545e-05,
+      "loss": 1.8703,
+      "step": 21452
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2813036859108883e-05,
+      "loss": 1.7732,
+      "step": 21453
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2812461501680116e-05,
+      "loss": 2.162,
+      "step": 21454
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2811886134141213e-05,
+      "loss": 1.7355,
+      "step": 21455
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2811310756494244e-05,
+      "loss": 1.9464,
+      "step": 21456
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.281073536874128e-05,
+      "loss": 2.0208,
+      "step": 21457
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2810159970884386e-05,
+      "loss": 2.0039,
+      "step": 21458
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.280958456292563e-05,
+      "loss": 2.0544,
+      "step": 21459
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2809009144867081e-05,
+      "loss": 1.5395,
+      "step": 21460
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2808433716710812e-05,
+      "loss": 1.7453,
+      "step": 21461
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2807858278458886e-05,
+      "loss": 2.1524,
+      "step": 21462
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2807282830113375e-05,
+      "loss": 1.9605,
+      "step": 21463
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2806707371676346e-05,
+      "loss": 1.8866,
+      "step": 21464
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2806131903149868e-05,
+      "loss": 1.6806,
+      "step": 21465
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2805556424536006e-05,
+      "loss": 1.7438,
+      "step": 21466
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2804980935836837e-05,
+      "loss": 1.8865,
+      "step": 21467
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.280440543705442e-05,
+      "loss": 2.0733,
+      "step": 21468
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2803829928190832e-05,
+      "loss": 1.9994,
+      "step": 21469
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2803254409248137e-05,
+      "loss": 1.8267,
+      "step": 21470
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2802678880228406e-05,
+      "loss": 1.6148,
+      "step": 21471
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2802103341133707e-05,
+      "loss": 2.114,
+      "step": 21472
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2801527791966109e-05,
+      "loss": 2.0662,
+      "step": 21473
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2800952232727683e-05,
+      "loss": 1.4415,
+      "step": 21474
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2800376663420493e-05,
+      "loss": 1.6247,
+      "step": 21475
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2799801084046612e-05,
+      "loss": 2.1452,
+      "step": 21476
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2799225494608113e-05,
+      "loss": 1.9077,
+      "step": 21477
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2798649895107053e-05,
+      "loss": 2.123,
+      "step": 21478
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2798074285545512e-05,
+      "loss": 1.7818,
+      "step": 21479
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2797498665925553e-05,
+      "loss": 1.8913,
+      "step": 21480
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2796923036249253e-05,
+      "loss": 2.1437,
+      "step": 21481
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2796347396518674e-05,
+      "loss": 2.0159,
+      "step": 21482
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2795771746735885e-05,
+      "loss": 1.7868,
+      "step": 21483
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2795196086902963e-05,
+      "loss": 1.832,
+      "step": 21484
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2794620417021966e-05,
+      "loss": 1.615,
+      "step": 21485
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2794044737094974e-05,
+      "loss": 1.8563,
+      "step": 21486
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.279346904712405e-05,
+      "loss": 1.668,
+      "step": 21487
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2792893347111266e-05,
+      "loss": 1.9058,
+      "step": 21488
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2792317637058688e-05,
+      "loss": 1.6937,
+      "step": 21489
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2791741916968392e-05,
+      "loss": 1.9334,
+      "step": 21490
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2791166186842445e-05,
+      "loss": 2.0724,
+      "step": 21491
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2790590446682911e-05,
+      "loss": 1.6464,
+      "step": 21492
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2790014696491868e-05,
+      "loss": 2.1182,
+      "step": 21493
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.278943893627138e-05,
+      "loss": 2.0931,
+      "step": 21494
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2788863166023522e-05,
+      "loss": 1.9298,
+      "step": 21495
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2788287385750358e-05,
+      "loss": 1.767,
+      "step": 21496
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.278771159545396e-05,
+      "loss": 1.8531,
+      "step": 21497
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2787135795136401e-05,
+      "loss": 1.7799,
+      "step": 21498
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2786559984799742e-05,
+      "loss": 1.9508,
+      "step": 21499
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2785984164446063e-05,
+      "loss": 2.0335,
+      "step": 21500
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.278540833407743e-05,
+      "loss": 2.1066,
+      "step": 21501
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.278483249369591e-05,
+      "loss": 2.4721,
+      "step": 21502
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2784256643303578e-05,
+      "loss": 1.9116,
+      "step": 21503
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2783680782902503e-05,
+      "loss": 2.0081,
+      "step": 21504
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2783104912494752e-05,
+      "loss": 1.9499,
+      "step": 21505
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2782529032082395e-05,
+      "loss": 1.3915,
+      "step": 21506
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.278195314166751e-05,
+      "loss": 1.6967,
+      "step": 21507
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2781377241252158e-05,
+      "loss": 1.9118,
+      "step": 21508
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2780801330838412e-05,
+      "loss": 1.8627,
+      "step": 21509
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2780225410428344e-05,
+      "loss": 1.6649,
+      "step": 21510
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2779649480024023e-05,
+      "loss": 1.8962,
+      "step": 21511
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.277907353962752e-05,
+      "loss": 1.8321,
+      "step": 21512
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.27784975892409e-05,
+      "loss": 1.6633,
+      "step": 21513
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2777921628866245e-05,
+      "loss": 1.6819,
+      "step": 21514
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2777345658505616e-05,
+      "loss": 1.6726,
+      "step": 21515
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2776769678161086e-05,
+      "loss": 1.6073,
+      "step": 21516
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2776193687834727e-05,
+      "loss": 1.5915,
+      "step": 21517
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2775617687528605e-05,
+      "loss": 1.7442,
+      "step": 21518
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2775041677244798e-05,
+      "loss": 1.9215,
+      "step": 21519
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2774465656985366e-05,
+      "loss": 2.2697,
+      "step": 21520
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2773889626752393e-05,
+      "loss": 1.8329,
+      "step": 21521
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.277331358654794e-05,
+      "loss": 1.5996,
+      "step": 21522
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.277273753637408e-05,
+      "loss": 1.5976,
+      "step": 21523
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2772161476232885e-05,
+      "loss": 1.831,
+      "step": 21524
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2771585406126425e-05,
+      "loss": 1.8125,
+      "step": 21525
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2771009326056771e-05,
+      "loss": 1.7439,
+      "step": 21526
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2770433236025994e-05,
+      "loss": 1.8698,
+      "step": 21527
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2769857136036165e-05,
+      "loss": 1.7158,
+      "step": 21528
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2769281026089353e-05,
+      "loss": 1.9467,
+      "step": 21529
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2768704906187632e-05,
+      "loss": 1.8742,
+      "step": 21530
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2768128776333073e-05,
+      "loss": 1.8988,
+      "step": 21531
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2767552636527745e-05,
+      "loss": 1.8635,
+      "step": 21532
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2766976486773718e-05,
+      "loss": 1.9541,
+      "step": 21533
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2766400327073067e-05,
+      "loss": 2.1559,
+      "step": 21534
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.276582415742786e-05,
+      "loss": 1.7999,
+      "step": 21535
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2765247977840167e-05,
+      "loss": 1.747,
+      "step": 21536
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2764671788312064e-05,
+      "loss": 1.9073,
+      "step": 21537
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.276409558884562e-05,
+      "loss": 1.9541,
+      "step": 21538
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2763519379442906e-05,
+      "loss": 1.6735,
+      "step": 21539
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2762943160105992e-05,
+      "loss": 2.0656,
+      "step": 21540
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.276236693083695e-05,
+      "loss": 1.7938,
+      "step": 21541
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2761790691637856e-05,
+      "loss": 1.9926,
+      "step": 21542
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2761214442510775e-05,
+      "loss": 1.7916,
+      "step": 21543
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.276063818345778e-05,
+      "loss": 1.892,
+      "step": 21544
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2760061914480946e-05,
+      "loss": 1.9279,
+      "step": 21545
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.275948563558234e-05,
+      "loss": 2.3762,
+      "step": 21546
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2758909346764039e-05,
+      "loss": 1.8285,
+      "step": 21547
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2758333048028108e-05,
+      "loss": 1.7639,
+      "step": 21548
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2757756739376625e-05,
+      "loss": 1.8204,
+      "step": 21549
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2757180420811654e-05,
+      "loss": 1.7899,
+      "step": 21550
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2756604092335279e-05,
+      "loss": 1.8933,
+      "step": 21551
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2756027753949556e-05,
+      "loss": 2.0708,
+      "step": 21552
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.275545140565657e-05,
+      "loss": 2.1192,
+      "step": 21553
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2754875047458386e-05,
+      "loss": 1.9097,
+      "step": 21554
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.2754298679357079e-05,
+      "loss": 1.0337,
+      "step": 21555
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2753722301354719e-05,
+      "loss": 1.6153,
+      "step": 21556
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2753145913453376e-05,
+      "loss": 2.0045,
+      "step": 21557
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.275256951565513e-05,
+      "loss": 1.6062,
+      "step": 21558
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2751993107962042e-05,
+      "loss": 1.7244,
+      "step": 21559
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2751416690376193e-05,
+      "loss": 1.9572,
+      "step": 21560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2750840262899653e-05,
+      "loss": 1.9246,
+      "step": 21561
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2750263825534488e-05,
+      "loss": 1.7768,
+      "step": 21562
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2749687378282778e-05,
+      "loss": 1.7125,
+      "step": 21563
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.274911092114659e-05,
+      "loss": 1.8253,
+      "step": 21564
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2748534454128003e-05,
+      "loss": 1.7302,
+      "step": 21565
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.274795797722908e-05,
+      "loss": 1.9771,
+      "step": 21566
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.27473814904519e-05,
+      "loss": 2.172,
+      "step": 21567
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2746804993798536e-05,
+      "loss": 1.8112,
+      "step": 21568
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2746228487271053e-05,
+      "loss": 1.955,
+      "step": 21569
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2745651970871529e-05,
+      "loss": 1.705,
+      "step": 21570
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2745075444602035e-05,
+      "loss": 2.0196,
+      "step": 21571
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2744498908464646e-05,
+      "loss": 1.8572,
+      "step": 21572
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2743922362461434e-05,
+      "loss": 1.757,
+      "step": 21573
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2743345806594465e-05,
+      "loss": 1.7005,
+      "step": 21574
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2742769240865822e-05,
+      "loss": 1.8519,
+      "step": 21575
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.2742192665277569e-05,
+      "loss": 1.7236,
+      "step": 21576
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.274161607983178e-05,
+      "loss": 2.0712,
+      "step": 21577
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2741039484530532e-05,
+      "loss": 2.2618,
+      "step": 21578
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2740462879375896e-05,
+      "loss": 1.7222,
+      "step": 21579
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2739886264369945e-05,
+      "loss": 1.7448,
+      "step": 21580
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2739309639514747e-05,
+      "loss": 2.3263,
+      "step": 21581
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2738733004812383e-05,
+      "loss": 1.7982,
+      "step": 21582
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2738156360264918e-05,
+      "loss": 1.6905,
+      "step": 21583
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.273757970587443e-05,
+      "loss": 1.7045,
+      "step": 21584
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2737003041642991e-05,
+      "loss": 1.5626,
+      "step": 21585
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2736426367572672e-05,
+      "loss": 1.7364,
+      "step": 21586
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2735849683665547e-05,
+      "loss": 1.9336,
+      "step": 21587
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.273527298992369e-05,
+      "loss": 1.995,
+      "step": 21588
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2734696286349175e-05,
+      "loss": 1.6368,
+      "step": 21589
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.273411957294407e-05,
+      "loss": 1.9551,
+      "step": 21590
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2733542849710455e-05,
+      "loss": 1.7523,
+      "step": 21591
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2732966116650398e-05,
+      "loss": 1.7873,
+      "step": 21592
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2732389373765977e-05,
+      "loss": 1.6854,
+      "step": 21593
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2731812621059262e-05,
+      "loss": 1.5815,
+      "step": 21594
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2731235858532324e-05,
+      "loss": 1.8937,
+      "step": 21595
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.273065908618724e-05,
+      "loss": 1.8203,
+      "step": 21596
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.273008230402608e-05,
+      "loss": 1.6723,
+      "step": 21597
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2729505512050927e-05,
+      "loss": 2.1619,
+      "step": 21598
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.272892871026384e-05,
+      "loss": 1.798,
+      "step": 21599
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2728351898666903e-05,
+      "loss": 1.921,
+      "step": 21600
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2727775077262185e-05,
+      "loss": 1.6217,
+      "step": 21601
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2727198246051762e-05,
+      "loss": 1.9246,
+      "step": 21602
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2726621405037705e-05,
+      "loss": 2.0406,
+      "step": 21603
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2726044554222087e-05,
+      "loss": 1.7945,
+      "step": 21604
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2725467693606988e-05,
+      "loss": 2.0842,
+      "step": 21605
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2724890823194474e-05,
+      "loss": 1.7363,
+      "step": 21606
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2724313942986623e-05,
+      "loss": 2.0511,
+      "step": 21607
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2723737052985507e-05,
+      "loss": 1.7619,
+      "step": 21608
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.27231601531932e-05,
+      "loss": 1.8846,
+      "step": 21609
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2722583243611779e-05,
+      "loss": 1.7313,
+      "step": 21610
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.272200632424331e-05,
+      "loss": 2.1349,
+      "step": 21611
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.272142939508988e-05,
+      "loss": 1.8101,
+      "step": 21612
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2720852456153547e-05,
+      "loss": 1.8866,
+      "step": 21613
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2720275507436397e-05,
+      "loss": 1.6221,
+      "step": 21614
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2719698548940499e-05,
+      "loss": 1.728,
+      "step": 21615
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.2719121580667929e-05,
+      "loss": 1.8933,
+      "step": 21616
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.271854460262076e-05,
+      "loss": 1.6503,
+      "step": 21617
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2717967614801064e-05,
+      "loss": 1.8578,
+      "step": 21618
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2717390617210919e-05,
+      "loss": 1.9109,
+      "step": 21619
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2716813609852394e-05,
+      "loss": 1.7196,
+      "step": 21620
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2716236592727568e-05,
+      "loss": 1.8923,
+      "step": 21621
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2715659565838516e-05,
+      "loss": 1.4686,
+      "step": 21622
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2715082529187309e-05,
+      "loss": 1.9823,
+      "step": 21623
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2714505482776022e-05,
+      "loss": 1.9315,
+      "step": 21624
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2713928426606731e-05,
+      "loss": 1.9224,
+      "step": 21625
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.271335136068151e-05,
+      "loss": 2.1775,
+      "step": 21626
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2712774285002428e-05,
+      "loss": 1.8363,
+      "step": 21627
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2712197199571567e-05,
+      "loss": 1.6133,
+      "step": 21628
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2711620104390999e-05,
+      "loss": 1.7688,
+      "step": 21629
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2711042999462797e-05,
+      "loss": 1.85,
+      "step": 21630
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.2710465884789037e-05,
+      "loss": 2.8063,
+      "step": 21631
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2709888760371794e-05,
+      "loss": 1.7835,
+      "step": 21632
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.270931162621314e-05,
+      "loss": 1.7479,
+      "step": 21633
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.270873448231515e-05,
+      "loss": 1.9274,
+      "step": 21634
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2708157328679906e-05,
+      "loss": 1.8439,
+      "step": 21635
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2707580165309473e-05,
+      "loss": 1.7908,
+      "step": 21636
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2707002992205928e-05,
+      "loss": 1.9503,
+      "step": 21637
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.270642580937135e-05,
+      "loss": 2.0933,
+      "step": 21638
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2705848616807809e-05,
+      "loss": 2.0501,
+      "step": 21639
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2705271414517386e-05,
+      "loss": 1.8866,
+      "step": 21640
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2704694202502148e-05,
+      "loss": 1.806,
+      "step": 21641
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2704116980764179e-05,
+      "loss": 1.7659,
+      "step": 21642
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2703539749305543e-05,
+      "loss": 2.0219,
+      "step": 21643
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2702962508128324e-05,
+      "loss": 1.8683,
+      "step": 21644
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2702385257234595e-05,
+      "loss": 1.9599,
+      "step": 21645
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2701807996626428e-05,
+      "loss": 1.7879,
+      "step": 21646
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2701230726305902e-05,
+      "loss": 2.1522,
+      "step": 21647
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2700653446275087e-05,
+      "loss": 2.296,
+      "step": 21648
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2700076156536066e-05,
+      "loss": 1.8538,
+      "step": 21649
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2699498857090907e-05,
+      "loss": 2.3067,
+      "step": 21650
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2698921547941687e-05,
+      "loss": 1.858,
+      "step": 21651
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2698344229090486e-05,
+      "loss": 1.9534,
+      "step": 21652
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2697766900539374e-05,
+      "loss": 1.574,
+      "step": 21653
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.269718956229043e-05,
+      "loss": 1.7193,
+      "step": 21654
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2696612214345723e-05,
+      "loss": 2.0881,
+      "step": 21655
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2696034856707336e-05,
+      "loss": 1.7326,
+      "step": 21656
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2695457489377339e-05,
+      "loss": 1.9761,
+      "step": 21657
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2694880112357809e-05,
+      "loss": 1.9626,
+      "step": 21658
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2694302725650827e-05,
+      "loss": 2.1434,
+      "step": 21659
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.269372532925846e-05,
+      "loss": 1.7078,
+      "step": 21660
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2693147923182787e-05,
+      "loss": 1.7954,
+      "step": 21661
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.2692570507425887e-05,
+      "loss": 1.8742,
+      "step": 21662
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2691993081989832e-05,
+      "loss": 1.7583,
+      "step": 21663
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.26914156468767e-05,
+      "loss": 1.8599,
+      "step": 21664
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2690838202088562e-05,
+      "loss": 2.0784,
+      "step": 21665
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2690260747627498e-05,
+      "loss": 1.7629,
+      "step": 21666
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2689683283495581e-05,
+      "loss": 2.071,
+      "step": 21667
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.268910580969489e-05,
+      "loss": 1.6187,
+      "step": 21668
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.2688528326227503e-05,
+      "loss": 1.9394,
+      "step": 21669
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.268795083309549e-05,
+      "loss": 1.8237,
+      "step": 21670
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2687373330300929e-05,
+      "loss": 2.0193,
+      "step": 21671
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2686795817845896e-05,
+      "loss": 1.9169,
+      "step": 21672
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2686218295732471e-05,
+      "loss": 2.2107,
+      "step": 21673
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.268564076396272e-05,
+      "loss": 2.3093,
+      "step": 21674
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2685063222538729e-05,
+      "loss": 1.467,
+      "step": 21675
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2684485671462571e-05,
+      "loss": 1.7143,
+      "step": 21676
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2683908110736325e-05,
+      "loss": 1.6442,
+      "step": 21677
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.268333054036206e-05,
+      "loss": 2.0781,
+      "step": 21678
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2682752960341858e-05,
+      "loss": 2.1507,
+      "step": 21679
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2682175370677791e-05,
+      "loss": 1.9473,
+      "step": 21680
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.268159777137194e-05,
+      "loss": 2.042,
+      "step": 21681
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2681020162426381e-05,
+      "loss": 1.9393,
+      "step": 21682
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2680442543843184e-05,
+      "loss": 2.2292,
+      "step": 21683
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2679864915624434e-05,
+      "loss": 1.7905,
+      "step": 21684
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.26792872777722e-05,
+      "loss": 1.6564,
+      "step": 21685
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2678709630288565e-05,
+      "loss": 1.9689,
+      "step": 21686
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.26781319731756e-05,
+      "loss": 1.7633,
+      "step": 21687
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2677554306435385e-05,
+      "loss": 1.895,
+      "step": 21688
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2676976630069996e-05,
+      "loss": 1.9221,
+      "step": 21689
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2676398944081507e-05,
+      "loss": 2.0094,
+      "step": 21690
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2675821248472e-05,
+      "loss": 1.6649,
+      "step": 21691
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2675243543243547e-05,
+      "loss": 1.7455,
+      "step": 21692
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2674665828398224e-05,
+      "loss": 2.0475,
+      "step": 21693
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2674088103938114e-05,
+      "loss": 1.5424,
+      "step": 21694
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2673510369865285e-05,
+      "loss": 2.1494,
+      "step": 21695
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2672932626181825e-05,
+      "loss": 1.8587,
+      "step": 21696
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2672354872889799e-05,
+      "loss": 1.6679,
+      "step": 21697
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2671777109991287e-05,
+      "loss": 1.8167,
+      "step": 21698
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2671199337488374e-05,
+      "loss": 1.9819,
+      "step": 21699
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.267062155538313e-05,
+      "loss": 2.0576,
+      "step": 21700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2670043763677631e-05,
+      "loss": 1.7649,
+      "step": 21701
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2669465962373956e-05,
+      "loss": 1.7521,
+      "step": 21702
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2668888151474184e-05,
+      "loss": 1.9862,
+      "step": 21703
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2668310330980388e-05,
+      "loss": 1.6977,
+      "step": 21704
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.266773250089465e-05,
+      "loss": 1.5918,
+      "step": 21705
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2667154661219044e-05,
+      "loss": 1.8964,
+      "step": 21706
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2666576811955647e-05,
+      "loss": 2.1249,
+      "step": 21707
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2665998953106536e-05,
+      "loss": 2.0366,
+      "step": 21708
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.266542108467379e-05,
+      "loss": 2.0654,
+      "step": 21709
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2664843206659487e-05,
+      "loss": 2.153,
+      "step": 21710
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2664265319065701e-05,
+      "loss": 1.8349,
+      "step": 21711
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.266368742189451e-05,
+      "loss": 1.8019,
+      "step": 21712
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2663109515147995e-05,
+      "loss": 1.8832,
+      "step": 21713
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2662531598828232e-05,
+      "loss": 1.6975,
+      "step": 21714
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2661953672937297e-05,
+      "loss": 2.1957,
+      "step": 21715
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.266137573747727e-05,
+      "loss": 1.9587,
+      "step": 21716
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2660797792450223e-05,
+      "loss": 1.9367,
+      "step": 21717
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.2660219837858233e-05,
+      "loss": 2.6556,
+      "step": 21718
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2659641873703393e-05,
+      "loss": 1.9694,
+      "step": 21719
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.265906389998776e-05,
+      "loss": 1.7302,
+      "step": 21720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2658485916713425e-05,
+      "loss": 2.0121,
+      "step": 21721
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2657907923882461e-05,
+      "loss": 2.1966,
+      "step": 21722
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2657329921496948e-05,
+      "loss": 1.7082,
+      "step": 21723
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2656751909558963e-05,
+      "loss": 2.0608,
+      "step": 21724
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.265617388807058e-05,
+      "loss": 2.0161,
+      "step": 21725
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2655595857033881e-05,
+      "loss": 1.7475,
+      "step": 21726
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2655017816450944e-05,
+      "loss": 1.9855,
+      "step": 21727
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2654439766323847e-05,
+      "loss": 1.8605,
+      "step": 21728
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2653861706654666e-05,
+      "loss": 1.7059,
+      "step": 21729
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2653283637445478e-05,
+      "loss": 1.6912,
+      "step": 21730
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2652705558698363e-05,
+      "loss": 1.8134,
+      "step": 21731
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.2652127470415398e-05,
+      "loss": 2.4395,
+      "step": 21732
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2651549372598667e-05,
+      "loss": 1.7554,
+      "step": 21733
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2650971265250238e-05,
+      "loss": 1.8994,
+      "step": 21734
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2650393148372196e-05,
+      "loss": 1.6293,
+      "step": 21735
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.264981502196662e-05,
+      "loss": 1.6892,
+      "step": 21736
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2649236886035581e-05,
+      "loss": 1.8219,
+      "step": 21737
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2648658740581165e-05,
+      "loss": 1.8226,
+      "step": 21738
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2648080585605446e-05,
+      "loss": 1.8851,
+      "step": 21739
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2647502421110505e-05,
+      "loss": 2.0963,
+      "step": 21740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2646924247098418e-05,
+      "loss": 1.9432,
+      "step": 21741
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.264634606357126e-05,
+      "loss": 1.6125,
+      "step": 21742
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.264576787053112e-05,
+      "loss": 1.7901,
+      "step": 21743
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2645189667980067e-05,
+      "loss": 1.8629,
+      "step": 21744
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2644611455920184e-05,
+      "loss": 2.2425,
+      "step": 21745
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2644033234353547e-05,
+      "loss": 1.7692,
+      "step": 21746
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.264345500328224e-05,
+      "loss": 1.7993,
+      "step": 21747
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2642876762708332e-05,
+      "loss": 1.8118,
+      "step": 21748
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.264229851263391e-05,
+      "loss": 1.931,
+      "step": 21749
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.264172025306105e-05,
+      "loss": 1.7846,
+      "step": 21750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2641141983991828e-05,
+      "loss": 2.0252,
+      "step": 21751
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2640563705428326e-05,
+      "loss": 2.1812,
+      "step": 21752
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2639985417372622e-05,
+      "loss": 1.6002,
+      "step": 21753
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2639407119826797e-05,
+      "loss": 1.6691,
+      "step": 21754
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2638828812792924e-05,
+      "loss": 2.1104,
+      "step": 21755
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2638250496273087e-05,
+      "loss": 1.8687,
+      "step": 21756
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2637672170269366e-05,
+      "loss": 2.0098,
+      "step": 21757
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2637093834783833e-05,
+      "loss": 1.8594,
+      "step": 21758
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.2636515489818575e-05,
+      "loss": 1.3633,
+      "step": 21759
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2635937135375667e-05,
+      "loss": 1.7637,
+      "step": 21760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2635358771457184e-05,
+      "loss": 1.5758,
+      "step": 21761
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2634780398065214e-05,
+      "loss": 1.5585,
+      "step": 21762
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2634202015201831e-05,
+      "loss": 1.6872,
+      "step": 21763
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2633623622869113e-05,
+      "loss": 2.122,
+      "step": 21764
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.263304522106914e-05,
+      "loss": 1.9818,
+      "step": 21765
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2632466809803998e-05,
+      "loss": 2.2163,
+      "step": 21766
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2631888389075755e-05,
+      "loss": 1.9382,
+      "step": 21767
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2631309958886497e-05,
+      "loss": 1.8521,
+      "step": 21768
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2630731519238302e-05,
+      "loss": 1.9412,
+      "step": 21769
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2630153070133248e-05,
+      "loss": 2.0068,
+      "step": 21770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2629574611573418e-05,
+      "loss": 1.7545,
+      "step": 21771
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2628996143560888e-05,
+      "loss": 1.5893,
+      "step": 21772
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2628417666097739e-05,
+      "loss": 1.9904,
+      "step": 21773
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.262783917918605e-05,
+      "loss": 1.8747,
+      "step": 21774
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.26272606828279e-05,
+      "loss": 1.7831,
+      "step": 21775
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.262668217702537e-05,
+      "loss": 2.0795,
+      "step": 21776
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.2626103661780537e-05,
+      "loss": 2.0262,
+      "step": 21777
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2625525137095487e-05,
+      "loss": 1.7861,
+      "step": 21778
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2624946602972288e-05,
+      "loss": 1.9216,
+      "step": 21779
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.2624368059413032e-05,
+      "loss": 1.8595,
+      "step": 21780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2623789506419792e-05,
+      "loss": 1.4989,
+      "step": 21781
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2623210943994649e-05,
+      "loss": 1.683,
+      "step": 21782
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2622632372139684e-05,
+      "loss": 2.0592,
+      "step": 21783
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2622053790856976e-05,
+      "loss": 2.1843,
+      "step": 21784
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2621475200148603e-05,
+      "loss": 1.6827,
+      "step": 21785
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.262089660001665e-05,
+      "loss": 2.2662,
+      "step": 21786
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.262031799046319e-05,
+      "loss": 1.7781,
+      "step": 21787
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2619739371490304e-05,
+      "loss": 1.8571,
+      "step": 21788
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.261916074310008e-05,
+      "loss": 2.1247,
+      "step": 21789
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.261858210529459e-05,
+      "loss": 1.9401,
+      "step": 21790
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2618003458075918e-05,
+      "loss": 1.8034,
+      "step": 21791
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2617424801446143e-05,
+      "loss": 1.8547,
+      "step": 21792
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.2616846135407344e-05,
+      "loss": 1.9037,
+      "step": 21793
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2616267459961602e-05,
+      "loss": 1.8484,
+      "step": 21794
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2615688775110994e-05,
+      "loss": 1.9894,
+      "step": 21795
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2615110080857609e-05,
+      "loss": 2.0889,
+      "step": 21796
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.2614531377203518e-05,
+      "loss": 1.6128,
+      "step": 21797
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2613952664150806e-05,
+      "loss": 1.9801,
+      "step": 21798
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2613373941701554e-05,
+      "loss": 1.6702,
+      "step": 21799
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.261279520985784e-05,
+      "loss": 2.0126,
+      "step": 21800
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2612216468621745e-05,
+      "loss": 1.4236,
+      "step": 21801
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2611637717995346e-05,
+      "loss": 1.8129,
+      "step": 21802
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2611058957980733e-05,
+      "loss": 2.0306,
+      "step": 21803
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2610480188579977e-05,
+      "loss": 1.4828,
+      "step": 21804
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2609901409795162e-05,
+      "loss": 1.8655,
+      "step": 21805
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2609322621628367e-05,
+      "loss": 2.0864,
+      "step": 21806
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.260874382408168e-05,
+      "loss": 1.9843,
+      "step": 21807
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2608165017157172e-05,
+      "loss": 2.3353,
+      "step": 21808
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2607586200856927e-05,
+      "loss": 2.057,
+      "step": 21809
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.260700737518303e-05,
+      "loss": 1.6884,
+      "step": 21810
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2606428540137552e-05,
+      "loss": 1.758,
+      "step": 21811
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2605849695722583e-05,
+      "loss": 2.2122,
+      "step": 21812
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2605270841940205e-05,
+      "loss": 1.8853,
+      "step": 21813
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2604691978792487e-05,
+      "loss": 1.8485,
+      "step": 21814
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.260411310628152e-05,
+      "loss": 1.6376,
+      "step": 21815
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2603534224409382e-05,
+      "loss": 1.5368,
+      "step": 21816
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2602955333178158e-05,
+      "loss": 2.0245,
+      "step": 21817
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2602376432589922e-05,
+      "loss": 1.7854,
+      "step": 21818
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2601797522646757e-05,
+      "loss": 2.0266,
+      "step": 21819
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2601218603350748e-05,
+      "loss": 2.0332,
+      "step": 21820
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.260063967470397e-05,
+      "loss": 1.8112,
+      "step": 21821
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.260006073670851e-05,
+      "loss": 1.8436,
+      "step": 21822
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2599481789366446e-05,
+      "loss": 1.9325,
+      "step": 21823
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.259890283267986e-05,
+      "loss": 2.1725,
+      "step": 21824
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2598323866650832e-05,
+      "loss": 1.8861,
+      "step": 21825
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.2597744891281442e-05,
+      "loss": 2.0246,
+      "step": 21826
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2597165906573778e-05,
+      "loss": 1.8955,
+      "step": 21827
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2596586912529913e-05,
+      "loss": 1.7947,
+      "step": 21828
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2596007909151937e-05,
+      "loss": 1.7818,
+      "step": 21829
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2595428896441924e-05,
+      "loss": 1.8571,
+      "step": 21830
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2594849874401955e-05,
+      "loss": 1.793,
+      "step": 21831
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2594270843034117e-05,
+      "loss": 2.1799,
+      "step": 21832
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.375,
+      "learning_rate": 1.259369180234049e-05,
+      "loss": 2.1485,
+      "step": 21833
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2593112752323154e-05,
+      "loss": 1.789,
+      "step": 21834
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2592533692984189e-05,
+      "loss": 1.513,
+      "step": 21835
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.259195462432568e-05,
+      "loss": 1.7541,
+      "step": 21836
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2591375546349705e-05,
+      "loss": 1.8992,
+      "step": 21837
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.259079645905835e-05,
+      "loss": 1.9688,
+      "step": 21838
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.2590217362453694e-05,
+      "loss": 1.8986,
+      "step": 21839
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2589638256537819e-05,
+      "loss": 1.6825,
+      "step": 21840
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2589059141312808e-05,
+      "loss": 2.0659,
+      "step": 21841
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2588480016780738e-05,
+      "loss": 1.9416,
+      "step": 21842
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2587900882943695e-05,
+      "loss": 2.5385,
+      "step": 21843
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2587321739803765e-05,
+      "loss": 1.8594,
+      "step": 21844
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2586742587363021e-05,
+      "loss": 1.6564,
+      "step": 21845
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2586163425623551e-05,
+      "loss": 1.7729,
+      "step": 21846
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2585584254587434e-05,
+      "loss": 2.1487,
+      "step": 21847
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2585005074256755e-05,
+      "loss": 1.5758,
+      "step": 21848
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2584425884633591e-05,
+      "loss": 1.6712,
+      "step": 21849
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2583846685720031e-05,
+      "loss": 2.0132,
+      "step": 21850
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2583267477518151e-05,
+      "loss": 1.5891,
+      "step": 21851
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2582688260030036e-05,
+      "loss": 2.0704,
+      "step": 21852
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2582109033257768e-05,
+      "loss": 2.1167,
+      "step": 21853
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2581529797203426e-05,
+      "loss": 1.5874,
+      "step": 21854
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2580950551869098e-05,
+      "loss": 2.2116,
+      "step": 21855
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.258037129725686e-05,
+      "loss": 1.7845,
+      "step": 21856
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2579792033368801e-05,
+      "loss": 1.7116,
+      "step": 21857
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.2579212760206997e-05,
+      "loss": 1.9591,
+      "step": 21858
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2578633477773534e-05,
+      "loss": 1.9824,
+      "step": 21859
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2578054186070494e-05,
+      "loss": 1.7003,
+      "step": 21860
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2577474885099958e-05,
+      "loss": 1.9976,
+      "step": 21861
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2576895574864012e-05,
+      "loss": 1.8032,
+      "step": 21862
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2576316255364732e-05,
+      "loss": 2.0727,
+      "step": 21863
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2575736926604207e-05,
+      "loss": 2.1974,
+      "step": 21864
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2575157588584515e-05,
+      "loss": 1.6669,
+      "step": 21865
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.257457824130774e-05,
+      "loss": 1.7691,
+      "step": 21866
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.257399888477597e-05,
+      "loss": 2.2047,
+      "step": 21867
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2573419518991277e-05,
+      "loss": 2.2961,
+      "step": 21868
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2572840143955752e-05,
+      "loss": 1.6655,
+      "step": 21869
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2572260759671476e-05,
+      "loss": 1.8009,
+      "step": 21870
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2571681366140528e-05,
+      "loss": 1.8028,
+      "step": 21871
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2571101963364996e-05,
+      "loss": 2.2265,
+      "step": 21872
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2570522551346957e-05,
+      "loss": 1.9268,
+      "step": 21873
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2569943130088501e-05,
+      "loss": 1.8889,
+      "step": 21874
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2569363699591708e-05,
+      "loss": 1.5787,
+      "step": 21875
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2568784259858658e-05,
+      "loss": 2.0721,
+      "step": 21876
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2568204810891437e-05,
+      "loss": 1.7797,
+      "step": 21877
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2567625352692127e-05,
+      "loss": 1.8639,
+      "step": 21878
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.256704588526281e-05,
+      "loss": 1.7177,
+      "step": 21879
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2566466408605571e-05,
+      "loss": 1.6631,
+      "step": 21880
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2565886922722493e-05,
+      "loss": 2.0095,
+      "step": 21881
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2565307427615656e-05,
+      "loss": 2.217,
+      "step": 21882
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2564727923287148e-05,
+      "loss": 1.4037,
+      "step": 21883
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2564148409739048e-05,
+      "loss": 1.8746,
+      "step": 21884
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.2563568886973441e-05,
+      "loss": 2.3599,
+      "step": 21885
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.256298935499241e-05,
+      "loss": 1.9161,
+      "step": 21886
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2562409813798042e-05,
+      "loss": 2.0777,
+      "step": 21887
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.256183026339241e-05,
+      "loss": 1.6548,
+      "step": 21888
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.256125070377761e-05,
+      "loss": 1.7762,
+      "step": 21889
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2560671134955715e-05,
+      "loss": 2.0408,
+      "step": 21890
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2560091556928815e-05,
+      "loss": 1.7525,
+      "step": 21891
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2559511969698992e-05,
+      "loss": 1.7811,
+      "step": 21892
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2558932373268326e-05,
+      "loss": 1.6505,
+      "step": 21893
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2558352767638908e-05,
+      "loss": 1.8142,
+      "step": 21894
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2557773152812814e-05,
+      "loss": 1.5476,
+      "step": 21895
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2557193528792132e-05,
+      "loss": 1.971,
+      "step": 21896
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2556613895578945e-05,
+      "loss": 1.8008,
+      "step": 21897
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2556034253175328e-05,
+      "loss": 1.6417,
+      "step": 21898
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2555454601583378e-05,
+      "loss": 2.099,
+      "step": 21899
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2554874940805173e-05,
+      "loss": 2.0231,
+      "step": 21900
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2554295270842797e-05,
+      "loss": 1.6511,
+      "step": 21901
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2553715591698334e-05,
+      "loss": 1.9304,
+      "step": 21902
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2553135903373865e-05,
+      "loss": 2.0617,
+      "step": 21903
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.255255620587148e-05,
+      "loss": 1.8504,
+      "step": 21904
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2551976499193255e-05,
+      "loss": 1.7691,
+      "step": 21905
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2551396783341281e-05,
+      "loss": 1.8837,
+      "step": 21906
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2550817058317637e-05,
+      "loss": 1.8629,
+      "step": 21907
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.255023732412441e-05,
+      "loss": 1.8123,
+      "step": 21908
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2549657580763683e-05,
+      "loss": 1.8724,
+      "step": 21909
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2549077828237537e-05,
+      "loss": 1.7415,
+      "step": 21910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2548498066548065e-05,
+      "loss": 2.0092,
+      "step": 21911
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2547918295697338e-05,
+      "loss": 1.7032,
+      "step": 21912
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2547338515687453e-05,
+      "loss": 1.7003,
+      "step": 21913
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2546758726520488e-05,
+      "loss": 1.9078,
+      "step": 21914
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2546178928198524e-05,
+      "loss": 1.9107,
+      "step": 21915
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2545599120723653e-05,
+      "loss": 1.6774,
+      "step": 21916
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2545019304097953e-05,
+      "loss": 2.1779,
+      "step": 21917
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2544439478323511e-05,
+      "loss": 1.5897,
+      "step": 21918
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.254385964340241e-05,
+      "loss": 2.1129,
+      "step": 21919
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2543279799336734e-05,
+      "loss": 1.8332,
+      "step": 21920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2542699946128572e-05,
+      "loss": 2.4969,
+      "step": 21921
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2542120083780003e-05,
+      "loss": 1.6575,
+      "step": 21922
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2265625,
+      "learning_rate": 1.2541540212293116e-05,
+      "loss": 2.2437,
+      "step": 21923
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2540960331669989e-05,
+      "loss": 1.9274,
+      "step": 21924
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2540380441912714e-05,
+      "loss": 1.9617,
+      "step": 21925
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2539800543023368e-05,
+      "loss": 2.0999,
+      "step": 21926
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2539220635004044e-05,
+      "loss": 1.8245,
+      "step": 21927
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2538640717856821e-05,
+      "loss": 1.8856,
+      "step": 21928
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2538060791583784e-05,
+      "loss": 1.8932,
+      "step": 21929
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2537480856187019e-05,
+      "loss": 1.8821,
+      "step": 21930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.253690091166861e-05,
+      "loss": 1.713,
+      "step": 21931
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2536320958030644e-05,
+      "loss": 2.0485,
+      "step": 21932
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.25357409952752e-05,
+      "loss": 1.8096,
+      "step": 21933
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2535161023404372e-05,
+      "loss": 2.3097,
+      "step": 21934
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2534581042420237e-05,
+      "loss": 1.9913,
+      "step": 21935
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2534001052324883e-05,
+      "loss": 2.0802,
+      "step": 21936
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2533421053120394e-05,
+      "loss": 1.6029,
+      "step": 21937
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2532841044808856e-05,
+      "loss": 1.7897,
+      "step": 21938
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2532261027392355e-05,
+      "loss": 1.6646,
+      "step": 21939
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2531681000872968e-05,
+      "loss": 1.9586,
+      "step": 21940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2531100965252793e-05,
+      "loss": 2.0096,
+      "step": 21941
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.2530520920533905e-05,
+      "loss": 2.1849,
+      "step": 21942
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2529940866718397e-05,
+      "loss": 1.8828,
+      "step": 21943
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2529360803808347e-05,
+      "loss": 2.9667,
+      "step": 21944
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2528780731805842e-05,
+      "loss": 2.0026,
+      "step": 21945
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2528200650712972e-05,
+      "loss": 1.7138,
+      "step": 21946
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2527620560531815e-05,
+      "loss": 1.9542,
+      "step": 21947
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2527040461264463e-05,
+      "loss": 1.7631,
+      "step": 21948
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2526460352912994e-05,
+      "loss": 2.2299,
+      "step": 21949
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2525880235479501e-05,
+      "loss": 1.8875,
+      "step": 21950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2525300108966066e-05,
+      "loss": 2.0,
+      "step": 21951
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2524719973374773e-05,
+      "loss": 2.1023,
+      "step": 21952
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2524139828707709e-05,
+      "loss": 2.0029,
+      "step": 21953
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2523559674966958e-05,
+      "loss": 1.5405,
+      "step": 21954
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.252297951215461e-05,
+      "loss": 1.9899,
+      "step": 21955
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2522399340272743e-05,
+      "loss": 1.8926,
+      "step": 21956
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.252181915932345e-05,
+      "loss": 1.7724,
+      "step": 21957
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2521238969308813e-05,
+      "loss": 1.7764,
+      "step": 21958
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2520658770230919e-05,
+      "loss": 1.8915,
+      "step": 21959
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.252007856209185e-05,
+      "loss": 2.0327,
+      "step": 21960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2519498344893699e-05,
+      "loss": 1.9082,
+      "step": 21961
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2518918118638544e-05,
+      "loss": 2.1144,
+      "step": 21962
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2518337883328473e-05,
+      "loss": 2.1356,
+      "step": 21963
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2517757638965575e-05,
+      "loss": 1.667,
+      "step": 21964
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.2517177385551933e-05,
+      "loss": 1.8055,
+      "step": 21965
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2516597123089633e-05,
+      "loss": 1.7449,
+      "step": 21966
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2516016851580763e-05,
+      "loss": 1.8838,
+      "step": 21967
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2515436571027408e-05,
+      "loss": 2.1242,
+      "step": 21968
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2514856281431653e-05,
+      "loss": 1.9213,
+      "step": 21969
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.2514275982795581e-05,
+      "loss": 2.0741,
+      "step": 21970
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2513695675121286e-05,
+      "loss": 1.9177,
+      "step": 21971
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2513115358410846e-05,
+      "loss": 2.016,
+      "step": 21972
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.2512535032666352e-05,
+      "loss": 1.7747,
+      "step": 21973
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.251195469788989e-05,
+      "loss": 1.7851,
+      "step": 21974
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.2511374354083544e-05,
+      "loss": 2.2544,
+      "step": 21975
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2510794001249402e-05,
+      "loss": 1.7529,
+      "step": 21976
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2510213639389545e-05,
+      "loss": 1.8729,
+      "step": 21977
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.250963326850607e-05,
+      "loss": 2.1751,
+      "step": 21978
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2509052888601054e-05,
+      "loss": 2.0523,
+      "step": 21979
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2508472499676585e-05,
+      "loss": 1.7787,
+      "step": 21980
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2507892101734753e-05,
+      "loss": 1.8478,
+      "step": 21981
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.250731169477764e-05,
+      "loss": 1.5501,
+      "step": 21982
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2506731278807335e-05,
+      "loss": 2.0699,
+      "step": 21983
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2506150853825922e-05,
+      "loss": 2.0338,
+      "step": 21984
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2505570419835493e-05,
+      "loss": 1.9353,
+      "step": 21985
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.250498997683813e-05,
+      "loss": 1.8759,
+      "step": 21986
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2504409524835919e-05,
+      "loss": 1.9244,
+      "step": 21987
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2503829063830951e-05,
+      "loss": 1.7485,
+      "step": 21988
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2503248593825306e-05,
+      "loss": 1.7147,
+      "step": 21989
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2502668114821075e-05,
+      "loss": 1.9531,
+      "step": 21990
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2502087626820345e-05,
+      "loss": 1.7462,
+      "step": 21991
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2501507129825204e-05,
+      "loss": 1.6028,
+      "step": 21992
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2500926623837733e-05,
+      "loss": 1.8611,
+      "step": 21993
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.250034610886002e-05,
+      "loss": 2.8679,
+      "step": 21994
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2499765584894159e-05,
+      "loss": 1.8134,
+      "step": 21995
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.2499185051942228e-05,
+      "loss": 2.006,
+      "step": 21996
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.249860451000632e-05,
+      "loss": 1.8983,
+      "step": 21997
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.249802395908852e-05,
+      "loss": 1.4643,
+      "step": 21998
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2497443399190917e-05,
+      "loss": 1.8106,
+      "step": 21999
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2496862830315593e-05,
+      "loss": 2.366,
+      "step": 22000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2496282252464636e-05,
+      "loss": 1.7095,
+      "step": 22001
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2495701665640136e-05,
+      "loss": 2.5476,
+      "step": 22002
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2495121069844178e-05,
+      "loss": 1.8391,
+      "step": 22003
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.249454046507885e-05,
+      "loss": 2.136,
+      "step": 22004
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.249395985134624e-05,
+      "loss": 1.8907,
+      "step": 22005
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2493379228648435e-05,
+      "loss": 1.8335,
+      "step": 22006
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2492798596987518e-05,
+      "loss": 1.9603,
+      "step": 22007
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2492217956365584e-05,
+      "loss": 1.5911,
+      "step": 22008
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2491637306784714e-05,
+      "loss": 2.0291,
+      "step": 22009
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2491056648246996e-05,
+      "loss": 1.803,
+      "step": 22010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2490475980754518e-05,
+      "loss": 1.7867,
+      "step": 22011
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.248989530430937e-05,
+      "loss": 1.865,
+      "step": 22012
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2489314618913635e-05,
+      "loss": 1.8556,
+      "step": 22013
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2488733924569405e-05,
+      "loss": 2.2266,
+      "step": 22014
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2488153221278763e-05,
+      "loss": 2.0233,
+      "step": 22015
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.24875725090438e-05,
+      "loss": 1.6795,
+      "step": 22016
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2486991787866601e-05,
+      "loss": 2.2336,
+      "step": 22017
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2486411057749258e-05,
+      "loss": 2.0288,
+      "step": 22018
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.248583031869385e-05,
+      "loss": 1.6633,
+      "step": 22019
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2485249570702471e-05,
+      "loss": 2.0341,
+      "step": 22020
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.248466881377721e-05,
+      "loss": 2.0669,
+      "step": 22021
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2484088047920154e-05,
+      "loss": 2.0623,
+      "step": 22022
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2483507273133384e-05,
+      "loss": 1.6648,
+      "step": 22023
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2482926489418994e-05,
+      "loss": 1.6672,
+      "step": 22024
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2482345696779073e-05,
+      "loss": 1.9561,
+      "step": 22025
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2481764895215703e-05,
+      "loss": 2.2183,
+      "step": 22026
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2481184084730977e-05,
+      "loss": 1.7709,
+      "step": 22027
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.248060326532698e-05,
+      "loss": 2.0668,
+      "step": 22028
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2480022437005803e-05,
+      "loss": 1.704,
+      "step": 22029
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.247944159976953e-05,
+      "loss": 1.7537,
+      "step": 22030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.247886075362025e-05,
+      "loss": 2.2711,
+      "step": 22031
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2478279898560055e-05,
+      "loss": 2.3491,
+      "step": 22032
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2477699034591026e-05,
+      "loss": 1.9261,
+      "step": 22033
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2477118161715258e-05,
+      "loss": 1.7941,
+      "step": 22034
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2476537279934836e-05,
+      "loss": 1.9924,
+      "step": 22035
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2475956389251846e-05,
+      "loss": 2.0796,
+      "step": 22036
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.247537548966838e-05,
+      "loss": 1.601,
+      "step": 22037
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2474794581186525e-05,
+      "loss": 1.941,
+      "step": 22038
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2474213663808368e-05,
+      "loss": 1.8987,
+      "step": 22039
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2473632737535995e-05,
+      "loss": 1.9048,
+      "step": 22040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.24730518023715e-05,
+      "loss": 2.0002,
+      "step": 22041
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.247247085831697e-05,
+      "loss": 1.7109,
+      "step": 22042
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.247188990537449e-05,
+      "loss": 1.9407,
+      "step": 22043
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2471308943546152e-05,
+      "loss": 1.6865,
+      "step": 22044
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2470727972834043e-05,
+      "loss": 1.5764,
+      "step": 22045
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.2470146993240248e-05,
+      "loss": 1.9129,
+      "step": 22046
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.246956600476686e-05,
+      "loss": 1.6068,
+      "step": 22047
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2468985007415968e-05,
+      "loss": 2.1054,
+      "step": 22048
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2468404001189658e-05,
+      "loss": 1.9111,
+      "step": 22049
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2467822986090019e-05,
+      "loss": 1.9475,
+      "step": 22050
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2467241962119141e-05,
+      "loss": 1.9758,
+      "step": 22051
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.246666092927911e-05,
+      "loss": 1.9878,
+      "step": 22052
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2466079887572019e-05,
+      "loss": 2.0205,
+      "step": 22053
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.246549883699995e-05,
+      "loss": 1.8961,
+      "step": 22054
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.2464917777565e-05,
+      "loss": 2.5231,
+      "step": 22055
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.246433670926925e-05,
+      "loss": 1.8887,
+      "step": 22056
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2463755632114794e-05,
+      "loss": 1.5706,
+      "step": 22057
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2463174546103718e-05,
+      "loss": 1.9015,
+      "step": 22058
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2462593451238114e-05,
+      "loss": 1.9322,
+      "step": 22059
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2462012347520068e-05,
+      "loss": 1.852,
+      "step": 22060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2461431234951666e-05,
+      "loss": 1.9492,
+      "step": 22061
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2460850113535008e-05,
+      "loss": 1.7842,
+      "step": 22062
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2460268983272172e-05,
+      "loss": 1.7939,
+      "step": 22063
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.2459687844165251e-05,
+      "loss": 1.6975,
+      "step": 22064
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2459106696216336e-05,
+      "loss": 1.6443,
+      "step": 22065
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2458525539427511e-05,
+      "loss": 1.9366,
+      "step": 22066
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.245794437380087e-05,
+      "loss": 1.8401,
+      "step": 22067
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2457363199338497e-05,
+      "loss": 1.7218,
+      "step": 22068
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2456782016042489e-05,
+      "loss": 1.7672,
+      "step": 22069
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2456200823914927e-05,
+      "loss": 1.9891,
+      "step": 22070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2455619622957906e-05,
+      "loss": 1.972,
+      "step": 22071
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2455038413173514e-05,
+      "loss": 2.0672,
+      "step": 22072
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2454457194563835e-05,
+      "loss": 1.8546,
+      "step": 22073
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2453875967130968e-05,
+      "loss": 2.1372,
+      "step": 22074
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2453294730876997e-05,
+      "loss": 1.8957,
+      "step": 22075
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.245271348580401e-05,
+      "loss": 1.7184,
+      "step": 22076
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2452132231914099e-05,
+      "loss": 1.9567,
+      "step": 22077
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2451550969209347e-05,
+      "loss": 1.7342,
+      "step": 22078
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2450969697691857e-05,
+      "loss": 1.7071,
+      "step": 22079
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2450388417363705e-05,
+      "loss": 1.8139,
+      "step": 22080
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.244980712822699e-05,
+      "loss": 1.9627,
+      "step": 22081
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2449225830283795e-05,
+      "loss": 1.8913,
+      "step": 22082
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2448644523536213e-05,
+      "loss": 1.6319,
+      "step": 22083
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2448063207986334e-05,
+      "loss": 1.7426,
+      "step": 22084
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2447481883636248e-05,
+      "loss": 1.7745,
+      "step": 22085
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2446900550488042e-05,
+      "loss": 2.1598,
+      "step": 22086
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2446319208543804e-05,
+      "loss": 1.8899,
+      "step": 22087
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2445737857805629e-05,
+      "loss": 1.4817,
+      "step": 22088
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2445156498275605e-05,
+      "loss": 1.7771,
+      "step": 22089
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2444575129955824e-05,
+      "loss": 1.9736,
+      "step": 22090
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2443993752848371e-05,
+      "loss": 1.7526,
+      "step": 22091
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.244341236695534e-05,
+      "loss": 1.8222,
+      "step": 22092
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.244283097227882e-05,
+      "loss": 1.892,
+      "step": 22093
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2442249568820896e-05,
+      "loss": 1.8628,
+      "step": 22094
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2441668156583667e-05,
+      "loss": 1.5805,
+      "step": 22095
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2441086735569217e-05,
+      "loss": 1.8364,
+      "step": 22096
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2440505305779635e-05,
+      "loss": 1.9349,
+      "step": 22097
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2439923867217018e-05,
+      "loss": 1.6457,
+      "step": 22098
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2439342419883449e-05,
+      "loss": 1.7289,
+      "step": 22099
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2438760963781022e-05,
+      "loss": 1.6312,
+      "step": 22100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2438179498911825e-05,
+      "loss": 2.1682,
+      "step": 22101
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.243759802527795e-05,
+      "loss": 1.9815,
+      "step": 22102
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2437016542881487e-05,
+      "loss": 2.1763,
+      "step": 22103
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2436435051724525e-05,
+      "loss": 1.857,
+      "step": 22104
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.2435853551809153e-05,
+      "loss": 1.7384,
+      "step": 22105
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2435272043137468e-05,
+      "loss": 2.0819,
+      "step": 22106
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2434690525711555e-05,
+      "loss": 1.9141,
+      "step": 22107
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2434108999533501e-05,
+      "loss": 2.0394,
+      "step": 22108
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2433527464605408e-05,
+      "loss": 1.8952,
+      "step": 22109
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2432945920929354e-05,
+      "loss": 1.7574,
+      "step": 22110
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2432364368507435e-05,
+      "loss": 2.157,
+      "step": 22111
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2431782807341743e-05,
+      "loss": 2.331,
+      "step": 22112
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2431201237434367e-05,
+      "loss": 1.9241,
+      "step": 22113
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2430619658787398e-05,
+      "loss": 1.8489,
+      "step": 22114
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2430038071402921e-05,
+      "loss": 2.3097,
+      "step": 22115
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.242945647528304e-05,
+      "loss": 1.92,
+      "step": 22116
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.242887487042983e-05,
+      "loss": 1.8732,
+      "step": 22117
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2428293256845394e-05,
+      "loss": 1.8786,
+      "step": 22118
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.2427711634531817e-05,
+      "loss": 1.8396,
+      "step": 22119
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.242713000349119e-05,
+      "loss": 2.0587,
+      "step": 22120
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2426548363725605e-05,
+      "loss": 1.8338,
+      "step": 22121
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2425966715237155e-05,
+      "loss": 1.6941,
+      "step": 22122
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2425385058027925e-05,
+      "loss": 2.594,
+      "step": 22123
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.242480339210001e-05,
+      "loss": 1.688,
+      "step": 22124
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.24242217174555e-05,
+      "loss": 2.0434,
+      "step": 22125
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2423640034096486e-05,
+      "loss": 1.9305,
+      "step": 22126
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.242305834202506e-05,
+      "loss": 1.965,
+      "step": 22127
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2422476641243312e-05,
+      "loss": 1.9488,
+      "step": 22128
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2421894931753336e-05,
+      "loss": 1.9603,
+      "step": 22129
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2421313213557218e-05,
+      "loss": 1.9035,
+      "step": 22130
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.242073148665705e-05,
+      "loss": 1.8028,
+      "step": 22131
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2420149751054928e-05,
+      "loss": 1.6627,
+      "step": 22132
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2419568006752938e-05,
+      "loss": 1.933,
+      "step": 22133
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2418986253753174e-05,
+      "loss": 1.6491,
+      "step": 22134
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2418404492057725e-05,
+      "loss": 1.9929,
+      "step": 22135
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2417822721668686e-05,
+      "loss": 1.8311,
+      "step": 22136
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2417240942588147e-05,
+      "loss": 1.9549,
+      "step": 22137
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2416659154818194e-05,
+      "loss": 1.9266,
+      "step": 22138
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2416077358360927e-05,
+      "loss": 1.6934,
+      "step": 22139
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.241549555321843e-05,
+      "loss": 1.6083,
+      "step": 22140
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2414913739392799e-05,
+      "loss": 1.9886,
+      "step": 22141
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2414331916886127e-05,
+      "loss": 1.9949,
+      "step": 22142
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.24137500857005e-05,
+      "loss": 1.9878,
+      "step": 22143
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2413168245838014e-05,
+      "loss": 1.7276,
+      "step": 22144
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2412586397300754e-05,
+      "loss": 1.8968,
+      "step": 22145
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2412004540090825e-05,
+      "loss": 1.5436,
+      "step": 22146
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2411422674210306e-05,
+      "loss": 1.6157,
+      "step": 22147
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2410840799661287e-05,
+      "loss": 1.7049,
+      "step": 22148
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.2410258916445875e-05,
+      "loss": 1.8902,
+      "step": 22149
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2409677024566145e-05,
+      "loss": 1.8894,
+      "step": 22150
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.24090951240242e-05,
+      "loss": 2.2131,
+      "step": 22151
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2408513214822125e-05,
+      "loss": 1.7809,
+      "step": 22152
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2407931296962016e-05,
+      "loss": 2.096,
+      "step": 22153
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2407349370445964e-05,
+      "loss": 1.9226,
+      "step": 22154
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.240676743527606e-05,
+      "loss": 1.8298,
+      "step": 22155
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2406185491454397e-05,
+      "loss": 1.974,
+      "step": 22156
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2405603538983065e-05,
+      "loss": 1.9016,
+      "step": 22157
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2405021577864155e-05,
+      "loss": 1.8835,
+      "step": 22158
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2404439608099765e-05,
+      "loss": 1.9006,
+      "step": 22159
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2403857629691981e-05,
+      "loss": 1.9639,
+      "step": 22160
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.24032756426429e-05,
+      "loss": 1.8804,
+      "step": 22161
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2402693646954607e-05,
+      "loss": 1.7896,
+      "step": 22162
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2402111642629204e-05,
+      "loss": 1.9229,
+      "step": 22163
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2401529629668774e-05,
+      "loss": 1.9404,
+      "step": 22164
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2400947608075414e-05,
+      "loss": 1.9313,
+      "step": 22165
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2400365577851215e-05,
+      "loss": 2.1855,
+      "step": 22166
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2399783538998269e-05,
+      "loss": 1.8261,
+      "step": 22167
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2399201491518668e-05,
+      "loss": 1.8694,
+      "step": 22168
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2398619435414507e-05,
+      "loss": 1.9646,
+      "step": 22169
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2398037370687876e-05,
+      "loss": 2.0383,
+      "step": 22170
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2397455297340865e-05,
+      "loss": 2.3825,
+      "step": 22171
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2396873215375573e-05,
+      "loss": 1.7439,
+      "step": 22172
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2396291124794086e-05,
+      "loss": 2.0397,
+      "step": 22173
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2395709025598501e-05,
+      "loss": 1.9646,
+      "step": 22174
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2395126917790909e-05,
+      "loss": 1.9865,
+      "step": 22175
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.23945448013734e-05,
+      "loss": 1.8648,
+      "step": 22176
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.239396267634807e-05,
+      "loss": 1.7117,
+      "step": 22177
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.239338054271701e-05,
+      "loss": 1.6558,
+      "step": 22178
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2392798400482313e-05,
+      "loss": 1.6915,
+      "step": 22179
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2392216249646072e-05,
+      "loss": 1.7403,
+      "step": 22180
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2391634090210377e-05,
+      "loss": 2.0888,
+      "step": 22181
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2391051922177328e-05,
+      "loss": 1.5871,
+      "step": 22182
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.239046974554901e-05,
+      "loss": 1.5642,
+      "step": 22183
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2389887560327518e-05,
+      "loss": 2.107,
+      "step": 22184
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2389305366514943e-05,
+      "loss": 1.7918,
+      "step": 22185
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2388723164113387e-05,
+      "loss": 1.8835,
+      "step": 22186
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.238814095312493e-05,
+      "loss": 1.6584,
+      "step": 22187
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2387558733551674e-05,
+      "loss": 1.6092,
+      "step": 22188
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.238697650539571e-05,
+      "loss": 1.7914,
+      "step": 22189
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2386394268659128e-05,
+      "loss": 1.8504,
+      "step": 22190
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2385812023344022e-05,
+      "loss": 2.8298,
+      "step": 22191
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2385229769452485e-05,
+      "loss": 2.0016,
+      "step": 22192
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2384647506986616e-05,
+      "loss": 1.8771,
+      "step": 22193
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.23840652359485e-05,
+      "loss": 2.0475,
+      "step": 22194
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2383482956340233e-05,
+      "loss": 1.6263,
+      "step": 22195
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.238290066816391e-05,
+      "loss": 2.1936,
+      "step": 22196
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2382318371421622e-05,
+      "loss": 1.9173,
+      "step": 22197
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2381736066115464e-05,
+      "loss": 1.9273,
+      "step": 22198
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2381153752247523e-05,
+      "loss": 1.7463,
+      "step": 22199
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2380571429819905e-05,
+      "loss": 1.8662,
+      "step": 22200
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2379989098834691e-05,
+      "loss": 1.5818,
+      "step": 22201
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.237940675929398e-05,
+      "loss": 1.9627,
+      "step": 22202
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2378824411199864e-05,
+      "loss": 1.6367,
+      "step": 22203
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2378242054554437e-05,
+      "loss": 1.7885,
+      "step": 22204
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2377659689359793e-05,
+      "loss": 2.1682,
+      "step": 22205
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2377077315618024e-05,
+      "loss": 1.7726,
+      "step": 22206
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5,
+      "learning_rate": 1.2376494933331225e-05,
+      "loss": 2.0151,
+      "step": 22207
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.2375912542501486e-05,
+      "loss": 1.7367,
+      "step": 22208
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2375330143130906e-05,
+      "loss": 1.8617,
+      "step": 22209
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2374747735221575e-05,
+      "loss": 1.9412,
+      "step": 22210
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2374165318775588e-05,
+      "loss": 1.8614,
+      "step": 22211
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.237358289379504e-05,
+      "loss": 1.9944,
+      "step": 22212
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2373000460282022e-05,
+      "loss": 1.7677,
+      "step": 22213
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2372418018238626e-05,
+      "loss": 1.8276,
+      "step": 22214
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2371835567666948e-05,
+      "loss": 1.8445,
+      "step": 22215
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2371253108569086e-05,
+      "loss": 1.9088,
+      "step": 22216
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2370670640947124e-05,
+      "loss": 1.6327,
+      "step": 22217
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2370088164803166e-05,
+      "loss": 2.2342,
+      "step": 22218
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2369505680139302e-05,
+      "loss": 2.055,
+      "step": 22219
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2368923186957624e-05,
+      "loss": 1.5735,
+      "step": 22220
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2368340685260228e-05,
+      "loss": 1.7913,
+      "step": 22221
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2367758175049205e-05,
+      "loss": 1.8618,
+      "step": 22222
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2367175656326655e-05,
+      "loss": 1.9006,
+      "step": 22223
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2366593129094669e-05,
+      "loss": 1.7349,
+      "step": 22224
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2366010593355335e-05,
+      "loss": 2.0423,
+      "step": 22225
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2365428049110757e-05,
+      "loss": 1.7709,
+      "step": 22226
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2364845496363022e-05,
+      "loss": 2.1407,
+      "step": 22227
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2364262935114229e-05,
+      "loss": 1.8048,
+      "step": 22228
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2363680365366466e-05,
+      "loss": 2.0296,
+      "step": 22229
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.2363097787121836e-05,
+      "loss": 2.2554,
+      "step": 22230
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2362515200382425e-05,
+      "loss": 1.8932,
+      "step": 22231
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2361932605150331e-05,
+      "loss": 1.7086,
+      "step": 22232
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.236135000142765e-05,
+      "loss": 1.8275,
+      "step": 22233
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2360767389216472e-05,
+      "loss": 2.009,
+      "step": 22234
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2360184768518895e-05,
+      "loss": 1.7869,
+      "step": 22235
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2359602139337008e-05,
+      "loss": 1.8998,
+      "step": 22236
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2359019501672915e-05,
+      "loss": 1.681,
+      "step": 22237
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2358436855528703e-05,
+      "loss": 1.8014,
+      "step": 22238
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2357854200906466e-05,
+      "loss": 2.1523,
+      "step": 22239
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2357271537808305e-05,
+      "loss": 1.9186,
+      "step": 22240
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2356688866236306e-05,
+      "loss": 1.751,
+      "step": 22241
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.235610618619257e-05,
+      "loss": 1.7644,
+      "step": 22242
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2355523497679187e-05,
+      "loss": 1.9168,
+      "step": 22243
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.2354940800698257e-05,
+      "loss": 2.1091,
+      "step": 22244
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.235435809525187e-05,
+      "loss": 2.0137,
+      "step": 22245
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.235377538134212e-05,
+      "loss": 1.7119,
+      "step": 22246
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2353192658971111e-05,
+      "loss": 1.6114,
+      "step": 22247
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2352609928140927e-05,
+      "loss": 2.087,
+      "step": 22248
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2352027188853666e-05,
+      "loss": 2.0223,
+      "step": 22249
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2351444441111427e-05,
+      "loss": 1.6821,
+      "step": 22250
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2350861684916298e-05,
+      "loss": 2.1468,
+      "step": 22251
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2350278920270378e-05,
+      "loss": 2.3649,
+      "step": 22252
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.234969614717576e-05,
+      "loss": 0.7544,
+      "step": 22253
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2349113365634542e-05,
+      "loss": 1.6174,
+      "step": 22254
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2348530575648815e-05,
+      "loss": 1.6375,
+      "step": 22255
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2347947777220677e-05,
+      "loss": 1.9461,
+      "step": 22256
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2347364970352223e-05,
+      "loss": 2.1208,
+      "step": 22257
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2346782155045545e-05,
+      "loss": 1.7332,
+      "step": 22258
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2346199331302742e-05,
+      "loss": 1.706,
+      "step": 22259
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.2345616499125907e-05,
+      "loss": 2.1417,
+      "step": 22260
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2345033658517137e-05,
+      "loss": 1.7706,
+      "step": 22261
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.234445080947852e-05,
+      "loss": 1.993,
+      "step": 22262
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2343867952012161e-05,
+      "loss": 2.6129,
+      "step": 22263
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2343285086120152e-05,
+      "loss": 1.9274,
+      "step": 22264
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2342702211804585e-05,
+      "loss": 1.8149,
+      "step": 22265
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2342119329067557e-05,
+      "loss": 2.0465,
+      "step": 22266
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2341536437911165e-05,
+      "loss": 1.7687,
+      "step": 22267
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2340953538337506e-05,
+      "loss": 1.9197,
+      "step": 22268
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2340370630348668e-05,
+      "loss": 1.9106,
+      "step": 22269
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2339787713946754e-05,
+      "loss": 1.7341,
+      "step": 22270
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2339204789133854e-05,
+      "loss": 2.22,
+      "step": 22271
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2338621855912068e-05,
+      "loss": 1.7141,
+      "step": 22272
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.233803891428349e-05,
+      "loss": 2.0613,
+      "step": 22273
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2337455964250213e-05,
+      "loss": 2.034,
+      "step": 22274
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2336873005814339e-05,
+      "loss": 1.9274,
+      "step": 22275
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.233629003897795e-05,
+      "loss": 2.0573,
+      "step": 22276
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2335707063743163e-05,
+      "loss": 2.0934,
+      "step": 22277
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2335124080112053e-05,
+      "loss": 1.7529,
+      "step": 22278
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2334541088086726e-05,
+      "loss": 1.7944,
+      "step": 22279
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2333958087669275e-05,
+      "loss": 1.9077,
+      "step": 22280
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.23333750788618e-05,
+      "loss": 1.9981,
+      "step": 22281
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2332792061666392e-05,
+      "loss": 1.6888,
+      "step": 22282
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2332209036085145e-05,
+      "loss": 1.639,
+      "step": 22283
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2331626002120163e-05,
+      "loss": 1.9613,
+      "step": 22284
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2331042959773533e-05,
+      "loss": 1.7301,
+      "step": 22285
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2330459909047356e-05,
+      "loss": 1.8141,
+      "step": 22286
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.232987684994373e-05,
+      "loss": 1.605,
+      "step": 22287
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2329293782464744e-05,
+      "loss": 1.9922,
+      "step": 22288
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.23287107066125e-05,
+      "loss": 1.7602,
+      "step": 22289
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.232812762238909e-05,
+      "loss": 2.0448,
+      "step": 22290
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2327544529796612e-05,
+      "loss": 1.856,
+      "step": 22291
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2326961428837161e-05,
+      "loss": 1.8872,
+      "step": 22292
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2326378319512834e-05,
+      "loss": 1.6928,
+      "step": 22293
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.232579520182573e-05,
+      "loss": 1.85,
+      "step": 22294
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2325212075777939e-05,
+      "loss": 1.8628,
+      "step": 22295
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2324628941371564e-05,
+      "loss": 1.9232,
+      "step": 22296
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2324045798608694e-05,
+      "loss": 2.1098,
+      "step": 22297
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.232346264749143e-05,
+      "loss": 1.657,
+      "step": 22298
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.2322879488021869e-05,
+      "loss": 1.7394,
+      "step": 22299
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2322296320202105e-05,
+      "loss": 2.1172,
+      "step": 22300
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2321713144034232e-05,
+      "loss": 1.739,
+      "step": 22301
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2321129959520353e-05,
+      "loss": 1.8978,
+      "step": 22302
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.232054676666256e-05,
+      "loss": 2.1351,
+      "step": 22303
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2319963565462949e-05,
+      "loss": 1.8003,
+      "step": 22304
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.231938035592362e-05,
+      "loss": 1.5079,
+      "step": 22305
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2318797138046663e-05,
+      "loss": 1.9473,
+      "step": 22306
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2318213911834182e-05,
+      "loss": 1.8311,
+      "step": 22307
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.231763067728827e-05,
+      "loss": 2.1891,
+      "step": 22308
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2317047434411023e-05,
+      "loss": 1.7928,
+      "step": 22309
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2316464183204537e-05,
+      "loss": 2.0593,
+      "step": 22310
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2315880923670913e-05,
+      "loss": 1.9992,
+      "step": 22311
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2315297655812245e-05,
+      "loss": 1.7662,
+      "step": 22312
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2314714379630623e-05,
+      "loss": 1.9839,
+      "step": 22313
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2314131095128159e-05,
+      "loss": 1.8639,
+      "step": 22314
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2313547802306938e-05,
+      "loss": 1.9142,
+      "step": 22315
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2312964501169056e-05,
+      "loss": 1.8985,
+      "step": 22316
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.2312381191716618e-05,
+      "loss": 1.7224,
+      "step": 22317
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2311797873951713e-05,
+      "loss": 2.0332,
+      "step": 22318
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2311214547876445e-05,
+      "loss": 1.8029,
+      "step": 22319
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2310631213492903e-05,
+      "loss": 1.6291,
+      "step": 22320
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2310047870803195e-05,
+      "loss": 1.8474,
+      "step": 22321
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2309464519809406e-05,
+      "loss": 1.9149,
+      "step": 22322
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2308881160513638e-05,
+      "loss": 1.8912,
+      "step": 22323
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2308297792917991e-05,
+      "loss": 1.6799,
+      "step": 22324
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2307714417024556e-05,
+      "loss": 1.8012,
+      "step": 22325
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.2307131032835435e-05,
+      "loss": 2.171,
+      "step": 22326
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2306547640352726e-05,
+      "loss": 1.7252,
+      "step": 22327
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2305964239578521e-05,
+      "loss": 1.7457,
+      "step": 22328
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2305380830514922e-05,
+      "loss": 1.8314,
+      "step": 22329
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.230479741316402e-05,
+      "loss": 1.965,
+      "step": 22330
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2304213987527922e-05,
+      "loss": 1.8608,
+      "step": 22331
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2303630553608714e-05,
+      "loss": 1.7748,
+      "step": 22332
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.2303047111408501e-05,
+      "loss": 1.8847,
+      "step": 22333
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.230246366092938e-05,
+      "loss": 1.9337,
+      "step": 22334
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2301880202173446e-05,
+      "loss": 1.8856,
+      "step": 22335
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2301296735142796e-05,
+      "loss": 1.6492,
+      "step": 22336
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.230071325983953e-05,
+      "loss": 2.179,
+      "step": 22337
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2300129776265744e-05,
+      "loss": 1.9451,
+      "step": 22338
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2299546284423532e-05,
+      "loss": 1.7017,
+      "step": 22339
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2298962784314999e-05,
+      "loss": 1.6392,
+      "step": 22340
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2298379275942236e-05,
+      "loss": 1.9081,
+      "step": 22341
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2297795759307345e-05,
+      "loss": 1.9831,
+      "step": 22342
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2297212234412419e-05,
+      "loss": 1.9861,
+      "step": 22343
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2296628701259561e-05,
+      "loss": 2.2307,
+      "step": 22344
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2296045159850864e-05,
+      "loss": 1.8597,
+      "step": 22345
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2295461610188428e-05,
+      "loss": 1.8772,
+      "step": 22346
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.229487805227435e-05,
+      "loss": 2.1083,
+      "step": 22347
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.229429448611073e-05,
+      "loss": 1.8503,
+      "step": 22348
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2293710911699664e-05,
+      "loss": 1.6219,
+      "step": 22349
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.2293127329043248e-05,
+      "loss": 2.4177,
+      "step": 22350
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2292543738143583e-05,
+      "loss": 1.9138,
+      "step": 22351
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2291960139002766e-05,
+      "loss": 1.9253,
+      "step": 22352
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2291376531622889e-05,
+      "loss": 1.8009,
+      "step": 22353
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2290792916006062e-05,
+      "loss": 1.9033,
+      "step": 22354
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2290209292154372e-05,
+      "loss": 2.1422,
+      "step": 22355
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2289625660069921e-05,
+      "loss": 1.8516,
+      "step": 22356
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2289042019754811e-05,
+      "loss": 1.7466,
+      "step": 22357
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2288458371211133e-05,
+      "loss": 1.6688,
+      "step": 22358
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.228787471444099e-05,
+      "loss": 1.5988,
+      "step": 22359
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2287291049446477e-05,
+      "loss": 2.0484,
+      "step": 22360
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2286707376229697e-05,
+      "loss": 1.6864,
+      "step": 22361
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.228612369479274e-05,
+      "loss": 1.8264,
+      "step": 22362
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2285540005137713e-05,
+      "loss": 1.6219,
+      "step": 22363
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2284956307266707e-05,
+      "loss": 1.7156,
+      "step": 22364
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2284372601181826e-05,
+      "loss": 1.7035,
+      "step": 22365
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2283788886885164e-05,
+      "loss": 1.6649,
+      "step": 22366
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.228320516437882e-05,
+      "loss": 1.9356,
+      "step": 22367
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2282621433664898e-05,
+      "loss": 1.9368,
+      "step": 22368
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2282037694745488e-05,
+      "loss": 1.7744,
+      "step": 22369
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2281453947622693e-05,
+      "loss": 2.0188,
+      "step": 22370
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2280870192298611e-05,
+      "loss": 2.0194,
+      "step": 22371
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2280286428775339e-05,
+      "loss": 1.6573,
+      "step": 22372
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2279702657054978e-05,
+      "loss": 1.7303,
+      "step": 22373
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2279118877139624e-05,
+      "loss": 1.8452,
+      "step": 22374
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.2278535089031377e-05,
+      "loss": 1.856,
+      "step": 22375
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2277951292732334e-05,
+      "loss": 2.118,
+      "step": 22376
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2277367488244595e-05,
+      "loss": 1.7167,
+      "step": 22377
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.227678367557026e-05,
+      "loss": 1.8462,
+      "step": 22378
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2276199854711428e-05,
+      "loss": 1.9007,
+      "step": 22379
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2275616025670194e-05,
+      "loss": 2.2468,
+      "step": 22380
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2275032188448657e-05,
+      "loss": 1.8377,
+      "step": 22381
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.227444834304892e-05,
+      "loss": 2.0509,
+      "step": 22382
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2273864489473075e-05,
+      "loss": 1.924,
+      "step": 22383
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.2273280627723229e-05,
+      "loss": 1.9297,
+      "step": 22384
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2272696757801476e-05,
+      "loss": 1.6907,
+      "step": 22385
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2272112879709911e-05,
+      "loss": 1.8582,
+      "step": 22386
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2271528993450643e-05,
+      "loss": 1.9187,
+      "step": 22387
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2270945099025764e-05,
+      "loss": 1.7975,
+      "step": 22388
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2270361196437374e-05,
+      "loss": 1.8615,
+      "step": 22389
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2269777285687572e-05,
+      "loss": 1.9698,
+      "step": 22390
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.226919336677846e-05,
+      "loss": 1.9929,
+      "step": 22391
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2268609439712134e-05,
+      "loss": 1.6382,
+      "step": 22392
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2268025504490689e-05,
+      "loss": 1.7668,
+      "step": 22393
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2267441561116233e-05,
+      "loss": 1.8265,
+      "step": 22394
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2266857609590859e-05,
+      "loss": 1.827,
+      "step": 22395
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.2266273649916669e-05,
+      "loss": 2.0975,
+      "step": 22396
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2265689682095761e-05,
+      "loss": 1.5556,
+      "step": 22397
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2265105706130235e-05,
+      "loss": 1.8154,
+      "step": 22398
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2264521722022189e-05,
+      "loss": 1.6538,
+      "step": 22399
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.375,
+      "learning_rate": 1.226393772977372e-05,
+      "loss": 2.0165,
+      "step": 22400
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.2263353729386937e-05,
+      "loss": 1.818,
+      "step": 22401
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2262769720863928e-05,
+      "loss": 1.921,
+      "step": 22402
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2262185704206797e-05,
+      "loss": 1.8318,
+      "step": 22403
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2261601679417643e-05,
+      "loss": 2.0121,
+      "step": 22404
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2261017646498568e-05,
+      "loss": 2.1261,
+      "step": 22405
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2260433605451669e-05,
+      "loss": 2.165,
+      "step": 22406
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2259849556279042e-05,
+      "loss": 2.0709,
+      "step": 22407
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2259265498982797e-05,
+      "loss": 1.7175,
+      "step": 22408
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.225868143356502e-05,
+      "loss": 1.9965,
+      "step": 22409
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2258097360027821e-05,
+      "loss": 1.9545,
+      "step": 22410
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2257513278373297e-05,
+      "loss": 2.084,
+      "step": 22411
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2256929188603546e-05,
+      "loss": 1.8965,
+      "step": 22412
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2256345090720667e-05,
+      "loss": 1.8185,
+      "step": 22413
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.225576098472676e-05,
+      "loss": 1.5474,
+      "step": 22414
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2255176870623928e-05,
+      "loss": 1.8385,
+      "step": 22415
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2254592748414268e-05,
+      "loss": 1.8496,
+      "step": 22416
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.225400861809988e-05,
+      "loss": 1.9707,
+      "step": 22417
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2253424479682863e-05,
+      "loss": 1.7753,
+      "step": 22418
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2252840333165319e-05,
+      "loss": 2.1096,
+      "step": 22419
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2252256178549347e-05,
+      "loss": 1.6705,
+      "step": 22420
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2251672015837046e-05,
+      "loss": 1.6434,
+      "step": 22421
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2251087845030519e-05,
+      "loss": 1.8379,
+      "step": 22422
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2250503666131859e-05,
+      "loss": 1.5891,
+      "step": 22423
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2249919479143173e-05,
+      "loss": 1.8046,
+      "step": 22424
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2249335284066558e-05,
+      "loss": 1.9252,
+      "step": 22425
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2248751080904118e-05,
+      "loss": 2.1382,
+      "step": 22426
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.2248166869657947e-05,
+      "loss": 1.6192,
+      "step": 22427
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2247582650330147e-05,
+      "loss": 2.0599,
+      "step": 22428
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.224699842292282e-05,
+      "loss": 1.7004,
+      "step": 22429
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2246414187438063e-05,
+      "loss": 1.9893,
+      "step": 22430
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.224582994387798e-05,
+      "loss": 2.2126,
+      "step": 22431
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.224524569224467e-05,
+      "loss": 1.5744,
+      "step": 22432
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2244661432540232e-05,
+      "loss": 1.7545,
+      "step": 22433
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2244077164766767e-05,
+      "loss": 2.0217,
+      "step": 22434
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.2243492888926376e-05,
+      "loss": 1.9951,
+      "step": 22435
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.224290860502116e-05,
+      "loss": 2.0623,
+      "step": 22436
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2242324313053213e-05,
+      "loss": 1.7003,
+      "step": 22437
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2241740013024646e-05,
+      "loss": 2.0901,
+      "step": 22438
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.224115570493755e-05,
+      "loss": 1.9116,
+      "step": 22439
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2240571388794032e-05,
+      "loss": 2.0559,
+      "step": 22440
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2239987064596189e-05,
+      "loss": 2.0173,
+      "step": 22441
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.223940273234612e-05,
+      "loss": 1.8896,
+      "step": 22442
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.223881839204593e-05,
+      "loss": 1.7774,
+      "step": 22443
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2238234043697714e-05,
+      "loss": 1.6688,
+      "step": 22444
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2237649687303584e-05,
+      "loss": 1.9493,
+      "step": 22445
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2237065322865625e-05,
+      "loss": 1.8651,
+      "step": 22446
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2236480950385947e-05,
+      "loss": 1.8724,
+      "step": 22447
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2235896569866648e-05,
+      "loss": 1.9027,
+      "step": 22448
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.223531218130983e-05,
+      "loss": 1.9291,
+      "step": 22449
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2234727784717598e-05,
+      "loss": 1.8587,
+      "step": 22450
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.223414338009204e-05,
+      "loss": 1.7388,
+      "step": 22451
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2233558967435273e-05,
+      "loss": 1.8962,
+      "step": 22452
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2232974546749383e-05,
+      "loss": 1.8441,
+      "step": 22453
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2232390118036482e-05,
+      "loss": 1.9884,
+      "step": 22454
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2231805681298665e-05,
+      "loss": 2.0531,
+      "step": 22455
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2231221236538036e-05,
+      "loss": 1.853,
+      "step": 22456
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2230636783756694e-05,
+      "loss": 1.7424,
+      "step": 22457
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2230052322956738e-05,
+      "loss": 2.1529,
+      "step": 22458
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.222946785414027e-05,
+      "loss": 1.9247,
+      "step": 22459
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2228883377309396e-05,
+      "loss": 1.9477,
+      "step": 22460
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.2228298892466211e-05,
+      "loss": 1.979,
+      "step": 22461
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.222771439961282e-05,
+      "loss": 1.6766,
+      "step": 22462
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2227129898751322e-05,
+      "loss": 2.1818,
+      "step": 22463
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.222654538988382e-05,
+      "loss": 1.8441,
+      "step": 22464
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.2225960873012412e-05,
+      "loss": 2.1229,
+      "step": 22465
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.2225376348139202e-05,
+      "loss": 1.9454,
+      "step": 22466
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2224791815266288e-05,
+      "loss": 2.0188,
+      "step": 22467
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.222420727439578e-05,
+      "loss": 1.8047,
+      "step": 22468
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2223622725529767e-05,
+      "loss": 1.601,
+      "step": 22469
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2223038168670356e-05,
+      "loss": 1.6943,
+      "step": 22470
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.222245360381965e-05,
+      "loss": 1.7325,
+      "step": 22471
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.222186903097975e-05,
+      "loss": 1.9006,
+      "step": 22472
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2221284450152755e-05,
+      "loss": 1.8089,
+      "step": 22473
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2220699861340767e-05,
+      "loss": 1.8253,
+      "step": 22474
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.222011526454589e-05,
+      "loss": 1.6388,
+      "step": 22475
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2219530659770225e-05,
+      "loss": 1.7875,
+      "step": 22476
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.221894604701587e-05,
+      "loss": 2.438,
+      "step": 22477
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2218361426284928e-05,
+      "loss": 1.9183,
+      "step": 22478
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.22177767975795e-05,
+      "loss": 1.9227,
+      "step": 22479
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2217192160901695e-05,
+      "loss": 1.8263,
+      "step": 22480
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2216607516253603e-05,
+      "loss": 2.2135,
+      "step": 22481
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2216022863637334e-05,
+      "loss": 1.7361,
+      "step": 22482
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2215438203054986e-05,
+      "loss": 2.2073,
+      "step": 22483
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.221485353450866e-05,
+      "loss": 2.0955,
+      "step": 22484
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2214268858000462e-05,
+      "loss": 1.7064,
+      "step": 22485
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2213684173532488e-05,
+      "loss": 1.9108,
+      "step": 22486
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2213099481106844e-05,
+      "loss": 1.6761,
+      "step": 22487
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2212514780725633e-05,
+      "loss": 1.7649,
+      "step": 22488
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2211930072390952e-05,
+      "loss": 2.478,
+      "step": 22489
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2211345356104907e-05,
+      "loss": 1.7012,
+      "step": 22490
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2210760631869596e-05,
+      "loss": 1.8029,
+      "step": 22491
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.221017589968713e-05,
+      "loss": 1.7418,
+      "step": 22492
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.2209591159559597e-05,
+      "loss": 1.8016,
+      "step": 22493
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2209006411489108e-05,
+      "loss": 1.7674,
+      "step": 22494
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2208421655477765e-05,
+      "loss": 1.9462,
+      "step": 22495
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.220783689152767e-05,
+      "loss": 1.7646,
+      "step": 22496
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.220725211964092e-05,
+      "loss": 1.7371,
+      "step": 22497
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.220666733981962e-05,
+      "loss": 1.983,
+      "step": 22498
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2206082552065877e-05,
+      "loss": 1.7818,
+      "step": 22499
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2205497756381784e-05,
+      "loss": 1.5152,
+      "step": 22500
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2204912952769453e-05,
+      "loss": 1.6051,
+      "step": 22501
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.2204328141230976e-05,
+      "loss": 3.0238,
+      "step": 22502
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2203743321768466e-05,
+      "loss": 1.8282,
+      "step": 22503
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2203158494384018e-05,
+      "loss": 1.6218,
+      "step": 22504
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2202573659079735e-05,
+      "loss": 1.8993,
+      "step": 22505
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2201988815857723e-05,
+      "loss": 1.6961,
+      "step": 22506
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2201403964720077e-05,
+      "loss": 1.8552,
+      "step": 22507
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2200819105668909e-05,
+      "loss": 1.8929,
+      "step": 22508
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2200234238706316e-05,
+      "loss": 1.8585,
+      "step": 22509
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2199649363834403e-05,
+      "loss": 1.7051,
+      "step": 22510
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2199064481055267e-05,
+      "loss": 1.8868,
+      "step": 22511
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2198479590371017e-05,
+      "loss": 1.714,
+      "step": 22512
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2197894691783752e-05,
+      "loss": 1.7317,
+      "step": 22513
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2197309785295572e-05,
+      "loss": 1.6769,
+      "step": 22514
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.2196724870908587e-05,
+      "loss": 1.9897,
+      "step": 22515
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2196139948624897e-05,
+      "loss": 1.9761,
+      "step": 22516
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2195555018446598e-05,
+      "loss": 1.8697,
+      "step": 22517
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2194970080375802e-05,
+      "loss": 2.0176,
+      "step": 22518
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2194385134414608e-05,
+      "loss": 1.7995,
+      "step": 22519
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2193800180565117e-05,
+      "loss": 1.6573,
+      "step": 22520
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.219321521882943e-05,
+      "loss": 1.7168,
+      "step": 22521
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2192630249209661e-05,
+      "loss": 2.0318,
+      "step": 22522
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2192045271707898e-05,
+      "loss": 1.8093,
+      "step": 22523
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2191460286326251e-05,
+      "loss": 1.8041,
+      "step": 22524
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2190875293066824e-05,
+      "loss": 1.7828,
+      "step": 22525
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.219029029193172e-05,
+      "loss": 1.8734,
+      "step": 22526
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.218970528292304e-05,
+      "loss": 1.9472,
+      "step": 22527
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.2189120266042882e-05,
+      "loss": 3.6444,
+      "step": 22528
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2188535241293362e-05,
+      "loss": 1.55,
+      "step": 22529
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.218795020867657e-05,
+      "loss": 1.8212,
+      "step": 22530
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2187365168194616e-05,
+      "loss": 2.0275,
+      "step": 22531
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2186780119849604e-05,
+      "loss": 1.8259,
+      "step": 22532
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2186195063643631e-05,
+      "loss": 2.0953,
+      "step": 22533
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2185609999578807e-05,
+      "loss": 1.9477,
+      "step": 22534
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.2185024927657228e-05,
+      "loss": 1.9896,
+      "step": 22535
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2184439847881005e-05,
+      "loss": 1.9296,
+      "step": 22536
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2183854760252234e-05,
+      "loss": 1.8831,
+      "step": 22537
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2183269664773023e-05,
+      "loss": 1.9735,
+      "step": 22538
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2182684561445475e-05,
+      "loss": 1.8545,
+      "step": 22539
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.2182099450271692e-05,
+      "loss": 1.9201,
+      "step": 22540
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2181514331253776e-05,
+      "loss": 2.0358,
+      "step": 22541
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2180929204393834e-05,
+      "loss": 2.0652,
+      "step": 22542
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2180344069693967e-05,
+      "loss": 1.9323,
+      "step": 22543
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2179758927156275e-05,
+      "loss": 1.6707,
+      "step": 22544
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2179173776782868e-05,
+      "loss": 1.9148,
+      "step": 22545
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2178588618575848e-05,
+      "loss": 1.8705,
+      "step": 22546
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2178003452537314e-05,
+      "loss": 1.9448,
+      "step": 22547
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2177418278669378e-05,
+      "loss": 1.935,
+      "step": 22548
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2176833096974134e-05,
+      "loss": 2.1301,
+      "step": 22549
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2176247907453692e-05,
+      "loss": 1.8311,
+      "step": 22550
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2175662710110148e-05,
+      "loss": 1.8388,
+      "step": 22551
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2175077504945619e-05,
+      "loss": 1.9257,
+      "step": 22552
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2174492291962197e-05,
+      "loss": 2.515,
+      "step": 22553
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2173907071161985e-05,
+      "loss": 1.8622,
+      "step": 22554
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2173321842547096e-05,
+      "loss": 1.9051,
+      "step": 22555
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2172736606119631e-05,
+      "loss": 2.0778,
+      "step": 22556
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.2172151361881687e-05,
+      "loss": 2.4881,
+      "step": 22557
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2171566109835373e-05,
+      "loss": 1.7533,
+      "step": 22558
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2170980849982797e-05,
+      "loss": 1.7676,
+      "step": 22559
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2170395582326055e-05,
+      "loss": 1.9833,
+      "step": 22560
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2169810306867254e-05,
+      "loss": 2.1218,
+      "step": 22561
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2169225023608498e-05,
+      "loss": 2.2426,
+      "step": 22562
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2168639732551893e-05,
+      "loss": 1.7076,
+      "step": 22563
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2168054433699538e-05,
+      "loss": 1.887,
+      "step": 22564
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2167469127053541e-05,
+      "loss": 1.9872,
+      "step": 22565
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2166883812616009e-05,
+      "loss": 2.1488,
+      "step": 22566
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2166298490389039e-05,
+      "loss": 1.7702,
+      "step": 22567
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.2165713160374735e-05,
+      "loss": 1.658,
+      "step": 22568
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2165127822575212e-05,
+      "loss": 2.1096,
+      "step": 22569
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2164542476992559e-05,
+      "loss": 1.9734,
+      "step": 22570
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2163957123628892e-05,
+      "loss": 1.9476,
+      "step": 22571
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.216337176248631e-05,
+      "loss": 1.6509,
+      "step": 22572
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2162786393566917e-05,
+      "loss": 2.0761,
+      "step": 22573
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.216220101687282e-05,
+      "loss": 2.2136,
+      "step": 22574
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2161615632406118e-05,
+      "loss": 2.0124,
+      "step": 22575
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2161030240168924e-05,
+      "loss": 2.0616,
+      "step": 22576
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.2160444840163334e-05,
+      "loss": 1.6951,
+      "step": 22577
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2159859432391456e-05,
+      "loss": 1.68,
+      "step": 22578
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2159274016855396e-05,
+      "loss": 1.6671,
+      "step": 22579
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2158688593557257e-05,
+      "loss": 1.9268,
+      "step": 22580
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2158103162499142e-05,
+      "loss": 1.5449,
+      "step": 22581
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2157517723683154e-05,
+      "loss": 1.7815,
+      "step": 22582
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2156932277111404e-05,
+      "loss": 1.7543,
+      "step": 22583
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2156346822785989e-05,
+      "loss": 2.3586,
+      "step": 22584
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.215576136070902e-05,
+      "loss": 1.7206,
+      "step": 22585
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2155175890882598e-05,
+      "loss": 1.6811,
+      "step": 22586
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2154590413308829e-05,
+      "loss": 1.9009,
+      "step": 22587
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2154004927989815e-05,
+      "loss": 1.817,
+      "step": 22588
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2153419434927664e-05,
+      "loss": 1.8661,
+      "step": 22589
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2152833934124479e-05,
+      "loss": 1.9628,
+      "step": 22590
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.2152248425582365e-05,
+      "loss": 1.6662,
+      "step": 22591
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2151662909303425e-05,
+      "loss": 2.2455,
+      "step": 22592
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2151077385289768e-05,
+      "loss": 1.9729,
+      "step": 22593
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2150491853543499e-05,
+      "loss": 1.7717,
+      "step": 22594
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2149906314066717e-05,
+      "loss": 2.0721,
+      "step": 22595
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2149320766861532e-05,
+      "loss": 1.9749,
+      "step": 22596
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.2148735211930046e-05,
+      "loss": 2.1716,
+      "step": 22597
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2148149649274363e-05,
+      "loss": 1.8848,
+      "step": 22598
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2147564078896594e-05,
+      "loss": 2.2284,
+      "step": 22599
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2146978500798835e-05,
+      "loss": 1.8866,
+      "step": 22600
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2146392914983203e-05,
+      "loss": 2.0609,
+      "step": 22601
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2145807321451791e-05,
+      "loss": 1.9373,
+      "step": 22602
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.214522172020671e-05,
+      "loss": 1.8398,
+      "step": 22603
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2144636111250065e-05,
+      "loss": 1.984,
+      "step": 22604
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2144050494583959e-05,
+      "loss": 1.8342,
+      "step": 22605
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2143464870210503e-05,
+      "loss": 1.8167,
+      "step": 22606
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2142879238131791e-05,
+      "loss": 2.0168,
+      "step": 22607
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.2142293598349939e-05,
+      "loss": 1.9086,
+      "step": 22608
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2141707950867047e-05,
+      "loss": 1.9348,
+      "step": 22609
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2141122295685224e-05,
+      "loss": 2.118,
+      "step": 22610
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.214053663280657e-05,
+      "loss": 1.7384,
+      "step": 22611
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2139950962233191e-05,
+      "loss": 1.7959,
+      "step": 22612
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.21393652839672e-05,
+      "loss": 1.8198,
+      "step": 22613
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2138779598010693e-05,
+      "loss": 1.9764,
+      "step": 22614
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2138193904365778e-05,
+      "loss": 1.9036,
+      "step": 22615
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2137608203034566e-05,
+      "loss": 1.989,
+      "step": 22616
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2137022494019155e-05,
+      "loss": 2.0309,
+      "step": 22617
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2136436777321654e-05,
+      "loss": 2.0,
+      "step": 22618
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2135851052944165e-05,
+      "loss": 2.0069,
+      "step": 22619
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.2135265320888803e-05,
+      "loss": 2.1251,
+      "step": 22620
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2134679581157662e-05,
+      "loss": 1.925,
+      "step": 22621
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2134093833752853e-05,
+      "loss": 1.9145,
+      "step": 22622
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2133508078676486e-05,
+      "loss": 1.7167,
+      "step": 22623
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.2132922315930657e-05,
+      "loss": 1.7466,
+      "step": 22624
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2132336545517481e-05,
+      "loss": 1.6644,
+      "step": 22625
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2131750767439056e-05,
+      "loss": 1.8436,
+      "step": 22626
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2131164981697493e-05,
+      "loss": 1.5105,
+      "step": 22627
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.375,
+      "learning_rate": 1.2130579188294894e-05,
+      "loss": 2.0628,
+      "step": 22628
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2129993387233365e-05,
+      "loss": 1.7322,
+      "step": 22629
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.212940757851502e-05,
+      "loss": 2.0881,
+      "step": 22630
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2128821762141954e-05,
+      "loss": 1.9804,
+      "step": 22631
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2128235938116278e-05,
+      "loss": 1.7296,
+      "step": 22632
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2127650106440097e-05,
+      "loss": 1.8454,
+      "step": 22633
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2127064267115515e-05,
+      "loss": 1.8851,
+      "step": 22634
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.212647842014464e-05,
+      "loss": 1.6714,
+      "step": 22635
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2125892565529584e-05,
+      "loss": 1.9686,
+      "step": 22636
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2125306703272444e-05,
+      "loss": 1.6378,
+      "step": 22637
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2124720833375326e-05,
+      "loss": 1.8676,
+      "step": 22638
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2124134955840341e-05,
+      "loss": 1.7903,
+      "step": 22639
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2123549070669594e-05,
+      "loss": 1.9741,
+      "step": 22640
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2122963177865189e-05,
+      "loss": 1.9105,
+      "step": 22641
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.2122377277429231e-05,
+      "loss": 5.4802,
+      "step": 22642
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2121791369363833e-05,
+      "loss": 1.4759,
+      "step": 22643
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2121205453671096e-05,
+      "loss": 2.0556,
+      "step": 22644
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2120619530353126e-05,
+      "loss": 1.8451,
+      "step": 22645
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2120033599412031e-05,
+      "loss": 1.8319,
+      "step": 22646
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2119447660849915e-05,
+      "loss": 1.64,
+      "step": 22647
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.2118861714668887e-05,
+      "loss": 2.0125,
+      "step": 22648
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2118275760871054e-05,
+      "loss": 1.9693,
+      "step": 22649
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.211768979945852e-05,
+      "loss": 1.701,
+      "step": 22650
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.211710383043339e-05,
+      "loss": 2.057,
+      "step": 22651
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.211651785379777e-05,
+      "loss": 1.8544,
+      "step": 22652
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2115931869553776e-05,
+      "loss": 2.0051,
+      "step": 22653
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.21153458777035e-05,
+      "loss": 2.3105,
+      "step": 22654
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.2114759878249063e-05,
+      "loss": 1.9876,
+      "step": 22655
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.211417387119256e-05,
+      "loss": 1.6343,
+      "step": 22656
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.2113587856536104e-05,
+      "loss": 1.903,
+      "step": 22657
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2113001834281796e-05,
+      "loss": 1.7635,
+      "step": 22658
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2112415804431748e-05,
+      "loss": 2.2515,
+      "step": 22659
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2111829766988069e-05,
+      "loss": 1.921,
+      "step": 22660
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2111243721952855e-05,
+      "loss": 1.7276,
+      "step": 22661
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2110657669328221e-05,
+      "loss": 1.987,
+      "step": 22662
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2110071609116272e-05,
+      "loss": 1.9144,
+      "step": 22663
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2109485541319118e-05,
+      "loss": 2.0119,
+      "step": 22664
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2108899465938858e-05,
+      "loss": 1.826,
+      "step": 22665
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2108313382977603e-05,
+      "loss": 1.7616,
+      "step": 22666
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2107727292437464e-05,
+      "loss": 1.8297,
+      "step": 22667
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2107141194320539e-05,
+      "loss": 1.6012,
+      "step": 22668
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2106555088628942e-05,
+      "loss": 2.0746,
+      "step": 22669
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.210596897536478e-05,
+      "loss": 1.8479,
+      "step": 22670
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2105382854530155e-05,
+      "loss": 1.8163,
+      "step": 22671
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2104796726127177e-05,
+      "loss": 1.8671,
+      "step": 22672
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2104210590157952e-05,
+      "loss": 1.6767,
+      "step": 22673
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2103624446624589e-05,
+      "loss": 1.957,
+      "step": 22674
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.210303829552919e-05,
+      "loss": 1.8584,
+      "step": 22675
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2102452136873868e-05,
+      "loss": 1.6244,
+      "step": 22676
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2101865970660729e-05,
+      "loss": 1.7232,
+      "step": 22677
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2101279796891878e-05,
+      "loss": 1.9453,
+      "step": 22678
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2100693615569422e-05,
+      "loss": 1.7081,
+      "step": 22679
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2100107426695473e-05,
+      "loss": 1.7469,
+      "step": 22680
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2099521230272131e-05,
+      "loss": 1.8691,
+      "step": 22681
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2098935026301506e-05,
+      "loss": 1.6384,
+      "step": 22682
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.209834881478571e-05,
+      "loss": 1.7691,
+      "step": 22683
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2097762595726841e-05,
+      "loss": 1.8623,
+      "step": 22684
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2097176369127015e-05,
+      "loss": 1.8651,
+      "step": 22685
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2096590134988335e-05,
+      "loss": 1.8421,
+      "step": 22686
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.209600389331291e-05,
+      "loss": 1.9769,
+      "step": 22687
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2095417644102846e-05,
+      "loss": 1.9205,
+      "step": 22688
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.209483138736025e-05,
+      "loss": 2.1337,
+      "step": 22689
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2094245123087233e-05,
+      "loss": 1.9264,
+      "step": 22690
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2093658851285899e-05,
+      "loss": 1.7618,
+      "step": 22691
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.2093072571958355e-05,
+      "loss": 1.7699,
+      "step": 22692
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2092486285106713e-05,
+      "loss": 1.9,
+      "step": 22693
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2091899990733075e-05,
+      "loss": 1.9959,
+      "step": 22694
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2091313688839554e-05,
+      "loss": 1.9649,
+      "step": 22695
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2090727379428251e-05,
+      "loss": 1.8781,
+      "step": 22696
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2090141062501284e-05,
+      "loss": 1.9158,
+      "step": 22697
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2089554738060747e-05,
+      "loss": 2.4607,
+      "step": 22698
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2088968406108759e-05,
+      "loss": 1.8085,
+      "step": 22699
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2088382066647422e-05,
+      "loss": 1.9903,
+      "step": 22700
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2087795719678846e-05,
+      "loss": 2.1291,
+      "step": 22701
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.208720936520514e-05,
+      "loss": 1.7767,
+      "step": 22702
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2086623003228403e-05,
+      "loss": 2.041,
+      "step": 22703
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.208603663375076e-05,
+      "loss": 1.8128,
+      "step": 22704
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.20854502567743e-05,
+      "loss": 1.424,
+      "step": 22705
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2084863872301145e-05,
+      "loss": 1.7669,
+      "step": 22706
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2084277480333395e-05,
+      "loss": 1.6912,
+      "step": 22707
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.208369108087316e-05,
+      "loss": 1.9481,
+      "step": 22708
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.208310467392255e-05,
+      "loss": 1.9949,
+      "step": 22709
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.208251825948367e-05,
+      "loss": 1.9281,
+      "step": 22710
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2081931837558631e-05,
+      "loss": 1.871,
+      "step": 22711
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2081345408149536e-05,
+      "loss": 2.0822,
+      "step": 22712
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.20807589712585e-05,
+      "loss": 1.8661,
+      "step": 22713
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.2080172526887627e-05,
+      "loss": 2.2407,
+      "step": 22714
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2079586075039025e-05,
+      "loss": 1.633,
+      "step": 22715
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2078999615714802e-05,
+      "loss": 1.848,
+      "step": 22716
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.2078413148917067e-05,
+      "loss": 1.888,
+      "step": 22717
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.207782667464793e-05,
+      "loss": 1.8317,
+      "step": 22718
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2077240192909497e-05,
+      "loss": 1.9423,
+      "step": 22719
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2076653703703877e-05,
+      "loss": 2.0778,
+      "step": 22720
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2076067207033178e-05,
+      "loss": 2.0408,
+      "step": 22721
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.2075480702899505e-05,
+      "loss": 1.5952,
+      "step": 22722
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2074894191304973e-05,
+      "loss": 1.846,
+      "step": 22723
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2074307672251688e-05,
+      "loss": 1.9294,
+      "step": 22724
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2073721145741756e-05,
+      "loss": 1.993,
+      "step": 22725
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2073134611777286e-05,
+      "loss": 1.8775,
+      "step": 22726
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.207254807036039e-05,
+      "loss": 1.8059,
+      "step": 22727
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2071961521493172e-05,
+      "loss": 1.852,
+      "step": 22728
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.207137496517774e-05,
+      "loss": 1.9105,
+      "step": 22729
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2070788401416209e-05,
+      "loss": 1.6241,
+      "step": 22730
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2070201830210681e-05,
+      "loss": 1.8631,
+      "step": 22731
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.206961525156327e-05,
+      "loss": 1.6372,
+      "step": 22732
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.206902866547608e-05,
+      "loss": 1.7694,
+      "step": 22733
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2068442071951218e-05,
+      "loss": 2.1409,
+      "step": 22734
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2067855470990802e-05,
+      "loss": 1.8032,
+      "step": 22735
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.2067268862596928e-05,
+      "loss": 2.0316,
+      "step": 22736
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2066682246771718e-05,
+      "loss": 2.453,
+      "step": 22737
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2066095623517267e-05,
+      "loss": 1.5967,
+      "step": 22738
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2065508992835695e-05,
+      "loss": 1.7987,
+      "step": 22739
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2064922354729108e-05,
+      "loss": 1.8211,
+      "step": 22740
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.206433570919961e-05,
+      "loss": 1.7148,
+      "step": 22741
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.2063749056249317e-05,
+      "loss": 1.8832,
+      "step": 22742
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.206316239588033e-05,
+      "loss": 1.9742,
+      "step": 22743
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2062575728094765e-05,
+      "loss": 2.0683,
+      "step": 22744
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2061989052894726e-05,
+      "loss": 1.9671,
+      "step": 22745
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2061402370282326e-05,
+      "loss": 1.9306,
+      "step": 22746
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2060815680259672e-05,
+      "loss": 2.0934,
+      "step": 22747
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2060228982828871e-05,
+      "loss": 2.0651,
+      "step": 22748
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2059642277992036e-05,
+      "loss": 1.6713,
+      "step": 22749
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.2059055565751272e-05,
+      "loss": 2.0499,
+      "step": 22750
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2058468846108694e-05,
+      "loss": 1.9374,
+      "step": 22751
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2057882119066404e-05,
+      "loss": 1.8689,
+      "step": 22752
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.2057295384626515e-05,
+      "loss": 1.9487,
+      "step": 22753
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2056708642791138e-05,
+      "loss": 1.5038,
+      "step": 22754
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2056121893562375e-05,
+      "loss": 1.6921,
+      "step": 22755
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2055535136942345e-05,
+      "loss": 1.7764,
+      "step": 22756
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.205494837293315e-05,
+      "loss": 1.9094,
+      "step": 22757
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2054361601536901e-05,
+      "loss": 1.5935,
+      "step": 22758
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2053774822755708e-05,
+      "loss": 2.0952,
+      "step": 22759
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.2053188036591682e-05,
+      "loss": 1.6562,
+      "step": 22760
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.2052601243046929e-05,
+      "loss": 1.9131,
+      "step": 22761
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.205201444212356e-05,
+      "loss": 2.1008,
+      "step": 22762
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2051427633823685e-05,
+      "loss": 1.9518,
+      "step": 22763
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2050840818149414e-05,
+      "loss": 2.2289,
+      "step": 22764
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2050253995102855e-05,
+      "loss": 1.6448,
+      "step": 22765
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2049667164686114e-05,
+      "loss": 1.6486,
+      "step": 22766
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2049080326901309e-05,
+      "loss": 2.1108,
+      "step": 22767
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.204849348175054e-05,
+      "loss": 1.5708,
+      "step": 22768
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2047906629235926e-05,
+      "loss": 1.6979,
+      "step": 22769
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.204731976935957e-05,
+      "loss": 1.9218,
+      "step": 22770
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2046732902123584e-05,
+      "loss": 2.08,
+      "step": 22771
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.2046146027530077e-05,
+      "loss": 2.0834,
+      "step": 22772
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.2045559145581157e-05,
+      "loss": 1.5751,
+      "step": 22773
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.204497225627894e-05,
+      "loss": 1.9489,
+      "step": 22774
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.2044385359625527e-05,
+      "loss": 1.7553,
+      "step": 22775
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2043798455623035e-05,
+      "loss": 1.7283,
+      "step": 22776
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.204321154427357e-05,
+      "loss": 1.8708,
+      "step": 22777
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2042624625579242e-05,
+      "loss": 2.1274,
+      "step": 22778
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2042037699542162e-05,
+      "loss": 1.8746,
+      "step": 22779
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2041450766164438e-05,
+      "loss": 1.9658,
+      "step": 22780
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2040863825448185e-05,
+      "loss": 1.9675,
+      "step": 22781
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2040276877395506e-05,
+      "loss": 1.9046,
+      "step": 22782
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.2039689922008515e-05,
+      "loss": 1.8077,
+      "step": 22783
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.2039102959289321e-05,
+      "loss": 1.8726,
+      "step": 22784
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2038515989240036e-05,
+      "loss": 1.8622,
+      "step": 22785
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2037929011862766e-05,
+      "loss": 1.8424,
+      "step": 22786
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2037342027159623e-05,
+      "loss": 1.9841,
+      "step": 22787
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.203675503513272e-05,
+      "loss": 2.728,
+      "step": 22788
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.203616803578416e-05,
+      "loss": 1.6782,
+      "step": 22789
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.203558102911606e-05,
+      "loss": 1.9146,
+      "step": 22790
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2034994015130529e-05,
+      "loss": 1.9099,
+      "step": 22791
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2034406993829672e-05,
+      "loss": 1.9681,
+      "step": 22792
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2033819965215606e-05,
+      "loss": 2.069,
+      "step": 22793
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.2033232929290437e-05,
+      "loss": 2.0776,
+      "step": 22794
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2032645886056276e-05,
+      "loss": 1.7036,
+      "step": 22795
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2032058835515232e-05,
+      "loss": 1.7748,
+      "step": 22796
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2031471777669421e-05,
+      "loss": 2.1713,
+      "step": 22797
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2030884712520949e-05,
+      "loss": 2.0587,
+      "step": 22798
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.375,
+      "learning_rate": 1.2030297640071925e-05,
+      "loss": 1.7955,
+      "step": 22799
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.2029710560324461e-05,
+      "loss": 1.8472,
+      "step": 22800
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.2029123473280668e-05,
+      "loss": 1.7696,
+      "step": 22801
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2028536378942658e-05,
+      "loss": 1.9792,
+      "step": 22802
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2027949277312537e-05,
+      "loss": 1.8118,
+      "step": 22803
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.202736216839242e-05,
+      "loss": 1.8609,
+      "step": 22804
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2026775052184414e-05,
+      "loss": 1.6124,
+      "step": 22805
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.202618792869063e-05,
+      "loss": 1.8207,
+      "step": 22806
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2025600797913182e-05,
+      "loss": 1.9454,
+      "step": 22807
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.2025013659854177e-05,
+      "loss": 1.9726,
+      "step": 22808
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2024426514515726e-05,
+      "loss": 1.9066,
+      "step": 22809
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2023839361899941e-05,
+      "loss": 2.0473,
+      "step": 22810
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.2023252202008933e-05,
+      "loss": 2.4505,
+      "step": 22811
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2022665034844811e-05,
+      "loss": 1.7211,
+      "step": 22812
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2022077860409685e-05,
+      "loss": 2.0161,
+      "step": 22813
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.202149067870567e-05,
+      "loss": 1.6562,
+      "step": 22814
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2020903489734872e-05,
+      "loss": 1.7254,
+      "step": 22815
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.2020316293499406e-05,
+      "loss": 1.8271,
+      "step": 22816
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.201972909000138e-05,
+      "loss": 1.7123,
+      "step": 22817
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2019141879242903e-05,
+      "loss": 1.7399,
+      "step": 22818
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.2018554661226092e-05,
+      "loss": 1.768,
+      "step": 22819
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2017967435953052e-05,
+      "loss": 1.8941,
+      "step": 22820
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.2017380203425899e-05,
+      "loss": 2.0138,
+      "step": 22821
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2016792963646737e-05,
+      "loss": 1.9156,
+      "step": 22822
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.2016205716617685e-05,
+      "loss": 1.9148,
+      "step": 22823
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.201561846234085e-05,
+      "loss": 1.5611,
+      "step": 22824
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.75,
+      "learning_rate": 1.2015031200818341e-05,
+      "loss": 1.9916,
+      "step": 22825
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2014443932052276e-05,
+      "loss": 2.02,
+      "step": 22826
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2013856656044756e-05,
+      "loss": 2.1257,
+      "step": 22827
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2013269372797902e-05,
+      "loss": 1.8281,
+      "step": 22828
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2012682082313815e-05,
+      "loss": 1.6971,
+      "step": 22829
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.2012094784594617e-05,
+      "loss": 1.766,
+      "step": 22830
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2011507479642412e-05,
+      "loss": 2.0852,
+      "step": 22831
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2010920167459315e-05,
+      "loss": 1.7739,
+      "step": 22832
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2010332848047435e-05,
+      "loss": 2.0784,
+      "step": 22833
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.200974552140888e-05,
+      "loss": 1.897,
+      "step": 22834
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2009158187545771e-05,
+      "loss": 2.0926,
+      "step": 22835
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.200857084646021e-05,
+      "loss": 1.7175,
+      "step": 22836
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.2007983498154313e-05,
+      "loss": 1.6861,
+      "step": 22837
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.200739614263019e-05,
+      "loss": 1.7389,
+      "step": 22838
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2006808779889952e-05,
+      "loss": 1.7722,
+      "step": 22839
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.2006221409935713e-05,
+      "loss": 1.9744,
+      "step": 22840
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2005634032769579e-05,
+      "loss": 2.0263,
+      "step": 22841
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.2005046648393667e-05,
+      "loss": 1.9976,
+      "step": 22842
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2004459256810086e-05,
+      "loss": 2.0444,
+      "step": 22843
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2003871858020948e-05,
+      "loss": 2.1006,
+      "step": 22844
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.2003284452028367e-05,
+      "loss": 2.0081,
+      "step": 22845
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.200269703883445e-05,
+      "loss": 1.7947,
+      "step": 22846
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.200210961844131e-05,
+      "loss": 1.8979,
+      "step": 22847
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2001522190851062e-05,
+      "loss": 2.0324,
+      "step": 22848
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.2000934756065814e-05,
+      "loss": 1.7729,
+      "step": 22849
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2000347314087677e-05,
+      "loss": 2.0591,
+      "step": 22850
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.1999759864918767e-05,
+      "loss": 1.8362,
+      "step": 22851
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.199917240856119e-05,
+      "loss": 2.159,
+      "step": 22852
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.1998584945017061e-05,
+      "loss": 1.7936,
+      "step": 22853
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1997997474288497e-05,
+      "loss": 1.9478,
+      "step": 22854
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.19974099963776e-05,
+      "loss": 1.8526,
+      "step": 22855
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1996822511286487e-05,
+      "loss": 1.9662,
+      "step": 22856
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.199623501901727e-05,
+      "loss": 1.5714,
+      "step": 22857
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.1995647519572062e-05,
+      "loss": 1.6309,
+      "step": 22858
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.199506001295297e-05,
+      "loss": 1.9641,
+      "step": 22859
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1994472499162111e-05,
+      "loss": 1.8796,
+      "step": 22860
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1993884978201595e-05,
+      "loss": 1.8931,
+      "step": 22861
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1993297450073532e-05,
+      "loss": 1.8594,
+      "step": 22862
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1992709914780039e-05,
+      "loss": 1.5589,
+      "step": 22863
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1992122372323222e-05,
+      "loss": 2.2388,
+      "step": 22864
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.19915348227052e-05,
+      "loss": 1.697,
+      "step": 22865
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1990947265928077e-05,
+      "loss": 2.1474,
+      "step": 22866
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.1990359701993972e-05,
+      "loss": 2.0588,
+      "step": 22867
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.1989772130904995e-05,
+      "loss": 1.6683,
+      "step": 22868
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.1989184552663258e-05,
+      "loss": 2.0298,
+      "step": 22869
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1988596967270871e-05,
+      "loss": 1.9479,
+      "step": 22870
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.1988009374729948e-05,
+      "loss": 1.6823,
+      "step": 22871
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.1987421775042605e-05,
+      "loss": 1.8364,
+      "step": 22872
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1986834168210946e-05,
+      "loss": 1.9571,
+      "step": 22873
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1986246554237091e-05,
+      "loss": 1.7039,
+      "step": 22874
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.198565893312315e-05,
+      "loss": 2.0573,
+      "step": 22875
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.1985071304871233e-05,
+      "loss": 1.981,
+      "step": 22876
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.1984483669483453e-05,
+      "loss": 1.9362,
+      "step": 22877
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.1983896026961927e-05,
+      "loss": 1.792,
+      "step": 22878
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1983308377308762e-05,
+      "loss": 1.8316,
+      "step": 22879
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.198272072052607e-05,
+      "loss": 1.8903,
+      "step": 22880
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.198213305661597e-05,
+      "loss": 1.9745,
+      "step": 22881
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.198154538558057e-05,
+      "loss": 2.3422,
+      "step": 22882
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.198095770742198e-05,
+      "loss": 1.8868,
+      "step": 22883
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.1980370022142318e-05,
+      "loss": 1.8419,
+      "step": 22884
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1979782329743691e-05,
+      "loss": 1.9079,
+      "step": 22885
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.197919463022822e-05,
+      "loss": 1.581,
+      "step": 22886
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1978606923598006e-05,
+      "loss": 2.3304,
+      "step": 22887
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.1978019209855174e-05,
+      "loss": 2.038,
+      "step": 22888
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.1977431489001829e-05,
+      "loss": 1.7587,
+      "step": 22889
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1976843761040082e-05,
+      "loss": 1.6352,
+      "step": 22890
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.375,
+      "learning_rate": 1.1976256025972053e-05,
+      "loss": 1.7869,
+      "step": 22891
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.1975668283799851e-05,
+      "loss": 1.9515,
+      "step": 22892
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.1975080534525588e-05,
+      "loss": 1.7541,
+      "step": 22893
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1974492778151376e-05,
+      "loss": 1.9267,
+      "step": 22894
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1973905014679333e-05,
+      "loss": 1.6529,
+      "step": 22895
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.1973317244111564e-05,
+      "loss": 1.5956,
+      "step": 22896
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1972729466450188e-05,
+      "loss": 1.9269,
+      "step": 22897
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1972141681697316e-05,
+      "loss": 2.0185,
+      "step": 22898
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1971553889855061e-05,
+      "loss": 1.7139,
+      "step": 22899
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.1970966090925536e-05,
+      "loss": 2.0648,
+      "step": 22900
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.1970378284910855e-05,
+      "loss": 1.6687,
+      "step": 22901
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.196979047181313e-05,
+      "loss": 1.6711,
+      "step": 22902
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.1969202651634474e-05,
+      "loss": 1.9647,
+      "step": 22903
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1968614824376995e-05,
+      "loss": 1.8285,
+      "step": 22904
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.196802699004282e-05,
+      "loss": 1.9756,
+      "step": 22905
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1967439148634048e-05,
+      "loss": 2.2509,
+      "step": 22906
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.19668513001528e-05,
+      "loss": 1.9145,
+      "step": 22907
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.1966263444601185e-05,
+      "loss": 1.5086,
+      "step": 22908
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.1965675581981318e-05,
+      "loss": 1.9049,
+      "step": 22909
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1965087712295314e-05,
+      "loss": 2.1182,
+      "step": 22910
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.196449983554528e-05,
+      "loss": 1.9816,
+      "step": 22911
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.1963911951733337e-05,
+      "loss": 1.7396,
+      "step": 22912
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.1963324060861594e-05,
+      "loss": 1.73,
+      "step": 22913
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.1962736162932168e-05,
+      "loss": 1.6985,
+      "step": 22914
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.1962148257947165e-05,
+      "loss": 1.5647,
+      "step": 22915
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1961560345908708e-05,
+      "loss": 1.9508,
+      "step": 22916
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1960972426818904e-05,
+      "loss": 2.0069,
+      "step": 22917
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.1960384500679865e-05,
+      "loss": 1.9713,
+      "step": 22918
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.195979656749371e-05,
+      "loss": 1.9097,
+      "step": 22919
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1959208627262549e-05,
+      "loss": 2.0732,
+      "step": 22920
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1958620679988498e-05,
+      "loss": 1.7945,
+      "step": 22921
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1958032725673667e-05,
+      "loss": 1.5218,
+      "step": 22922
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1957444764320172e-05,
+      "loss": 1.7424,
+      "step": 22923
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.1956856795930125e-05,
+      "loss": 1.8537,
+      "step": 22924
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.195626882050564e-05,
+      "loss": 2.1639,
+      "step": 22925
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1955680838048835e-05,
+      "loss": 1.7023,
+      "step": 22926
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.1955092848561816e-05,
+      "loss": 1.9897,
+      "step": 22927
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.1954504852046703e-05,
+      "loss": 1.5008,
+      "step": 22928
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1953916848505605e-05,
+      "loss": 1.7887,
+      "step": 22929
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.195332883794064e-05,
+      "loss": 1.83,
+      "step": 22930
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.195274082035392e-05,
+      "loss": 1.6952,
+      "step": 22931
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1952152795747558e-05,
+      "loss": 1.8942,
+      "step": 22932
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 52408,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 1638,
+  "total_flos": 4.2768779222596977e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}