diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,137312 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999816742267121,
+  "eval_steps": 500,
+  "global_step": 19610,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 5.0993456105390725e-05,
+      "grad_norm": 20.0,
+      "learning_rate": 0.00020360000000000002,
+      "loss": 6.9307,
+      "step": 1
+    },
+    {
+      "epoch": 0.00010198691221078145,
+      "grad_norm": 18.75,
+      "learning_rate": 0.00020720000000000002,
+      "loss": 6.7836,
+      "step": 2
+    },
+    {
+      "epoch": 0.00015298036831617217,
+      "grad_norm": 17.75,
+      "learning_rate": 0.00021080000000000003,
+      "loss": 6.6823,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002039738244215629,
+      "grad_norm": 16.5,
+      "learning_rate": 0.0002144,
+      "loss": 6.5402,
+      "step": 4
+    },
+    {
+      "epoch": 0.00025496728052695365,
+      "grad_norm": 15.25,
+      "learning_rate": 0.00021800000000000004,
+      "loss": 6.436,
+      "step": 5
+    },
+    {
+      "epoch": 0.00030596073663234435,
+      "grad_norm": 14.1875,
+      "learning_rate": 0.00022160000000000002,
+      "loss": 6.3405,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003569541927377351,
+      "grad_norm": 13.0625,
+      "learning_rate": 0.00022520000000000003,
+      "loss": 6.2694,
+      "step": 7
+    },
+    {
+      "epoch": 0.0004079476488431258,
+      "grad_norm": 12.1875,
+      "learning_rate": 0.0002288,
+      "loss": 6.1654,
+      "step": 8
+    },
+    {
+      "epoch": 0.00045894110494851655,
+      "grad_norm": 11.3125,
+      "learning_rate": 0.0002324,
+      "loss": 6.0625,
+      "step": 9
+    },
+    {
+      "epoch": 0.0005099345610539073,
+      "grad_norm": 10.4375,
+      "learning_rate": 0.00023600000000000002,
+      "loss": 6.0138,
+      "step": 10
+    },
+    {
+      "epoch": 0.000560928017159298,
+      "grad_norm": 9.9375,
+      "learning_rate": 0.00023960000000000002,
+      "loss": 5.9367,
+      "step": 11
+    },
+    {
+      "epoch": 0.0006119214732646887,
+      "grad_norm": 9.3125,
+      "learning_rate": 0.00024320000000000003,
+      "loss": 5.8797,
+      "step": 12
+    },
+    {
+      "epoch": 0.0006629149293700794,
+      "grad_norm": 8.875,
+      "learning_rate": 0.00024680000000000004,
+      "loss": 5.8179,
+      "step": 13
+    },
+    {
+      "epoch": 0.0007139083854754702,
+      "grad_norm": 8.625,
+      "learning_rate": 0.0002504,
+      "loss": 5.7682,
+      "step": 14
+    },
+    {
+      "epoch": 0.0007649018415808609,
+      "grad_norm": 8.375,
+      "learning_rate": 0.000254,
+      "loss": 5.6787,
+      "step": 15
+    },
+    {
+      "epoch": 0.0008158952976862516,
+      "grad_norm": 8.125,
+      "learning_rate": 0.00025760000000000003,
+      "loss": 5.6361,
+      "step": 16
+    },
+    {
+      "epoch": 0.0008668887537916423,
+      "grad_norm": 8.0625,
+      "learning_rate": 0.0002612,
+      "loss": 5.5871,
+      "step": 17
+    },
+    {
+      "epoch": 0.0009178822098970331,
+      "grad_norm": 7.90625,
+      "learning_rate": 0.00026480000000000004,
+      "loss": 5.4992,
+      "step": 18
+    },
+    {
+      "epoch": 0.0009688756660024238,
+      "grad_norm": 7.875,
+      "learning_rate": 0.0002684,
+      "loss": 5.399,
+      "step": 19
+    },
+    {
+      "epoch": 0.0010198691221078146,
+      "grad_norm": 7.84375,
+      "learning_rate": 0.00027200000000000005,
+      "loss": 5.3451,
+      "step": 20
+    },
+    {
+      "epoch": 0.0010708625782132053,
+      "grad_norm": 7.8125,
+      "learning_rate": 0.00027560000000000003,
+      "loss": 5.3446,
+      "step": 21
+    },
+    {
+      "epoch": 0.001121856034318596,
+      "grad_norm": 7.6875,
+      "learning_rate": 0.0002792,
+      "loss": 5.2659,
+      "step": 22
+    },
+    {
+      "epoch": 0.0011728494904239867,
+      "grad_norm": 7.375,
+      "learning_rate": 0.0002828,
+      "loss": 5.2222,
+      "step": 23
+    },
+    {
+      "epoch": 0.0012238429465293774,
+      "grad_norm": 7.125,
+      "learning_rate": 0.00028639999999999997,
+      "loss": 5.1494,
+      "step": 24
+    },
+    {
+      "epoch": 0.001274836402634768,
+      "grad_norm": 6.8125,
+      "learning_rate": 0.00029000000000000006,
+      "loss": 5.1262,
+      "step": 25
+    },
+    {
+      "epoch": 0.0013258298587401588,
+      "grad_norm": 6.40625,
+      "learning_rate": 0.00029360000000000003,
+      "loss": 5.0426,
+      "step": 26
+    },
+    {
+      "epoch": 0.0013768233148455495,
+      "grad_norm": 5.96875,
+      "learning_rate": 0.0002972,
+      "loss": 4.9856,
+      "step": 27
+    },
+    {
+      "epoch": 0.0014278167709509404,
+      "grad_norm": 5.5,
+      "learning_rate": 0.0003008,
+      "loss": 4.9369,
+      "step": 28
+    },
+    {
+      "epoch": 0.001478810227056331,
+      "grad_norm": 5.09375,
+      "learning_rate": 0.0003044,
+      "loss": 4.9122,
+      "step": 29
+    },
+    {
+      "epoch": 0.0015298036831617218,
+      "grad_norm": 4.6875,
+      "learning_rate": 0.000308,
+      "loss": 4.8626,
+      "step": 30
+    },
+    {
+      "epoch": 0.0015807971392671125,
+      "grad_norm": 4.59375,
+      "learning_rate": 0.0003116,
+      "loss": 4.8319,
+      "step": 31
+    },
+    {
+      "epoch": 0.0016317905953725032,
+      "grad_norm": 4.375,
+      "learning_rate": 0.0003152,
+      "loss": 4.7532,
+      "step": 32
+    },
+    {
+      "epoch": 0.0016827840514778939,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.00031880000000000005,
+      "loss": 4.7697,
+      "step": 33
+    },
+    {
+      "epoch": 0.0017337775075832846,
+      "grad_norm": 4.59375,
+      "learning_rate": 0.00032240000000000003,
+      "loss": 4.7271,
+      "step": 34
+    },
+    {
+      "epoch": 0.0017847709636886755,
+      "grad_norm": 4.65625,
+      "learning_rate": 0.00032600000000000006,
+      "loss": 4.6813,
+      "step": 35
+    },
+    {
+      "epoch": 0.0018357644197940662,
+      "grad_norm": 4.21875,
+      "learning_rate": 0.0003296,
+      "loss": 4.6441,
+      "step": 36
+    },
+    {
+      "epoch": 0.0018867578758994569,
+      "grad_norm": 3.875,
+      "learning_rate": 0.0003332,
+      "loss": 4.6106,
+      "step": 37
+    },
+    {
+      "epoch": 0.0019377513320048476,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.0003368,
+      "loss": 4.6054,
+      "step": 38
+    },
+    {
+      "epoch": 0.0019887447881102383,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00034040000000000003,
+      "loss": 4.549,
+      "step": 39
+    },
+    {
+      "epoch": 0.002039738244215629,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.000344,
+      "loss": 4.5294,
+      "step": 40
+    },
+    {
+      "epoch": 0.0020907317003210197,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00034760000000000005,
+      "loss": 4.4905,
+      "step": 41
+    },
+    {
+      "epoch": 0.0021417251564264106,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0003512,
+      "loss": 4.4439,
+      "step": 42
+    },
+    {
+      "epoch": 0.002192718612531801,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0003548,
+      "loss": 4.4197,
+      "step": 43
+    },
+    {
+      "epoch": 0.002243712068637192,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0003584,
+      "loss": 4.3697,
+      "step": 44
+    },
+    {
+      "epoch": 0.0022947055247425825,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.000362,
+      "loss": 4.3525,
+      "step": 45
+    },
+    {
+      "epoch": 0.0023456989808479734,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00036560000000000005,
+      "loss": 4.3537,
+      "step": 46
+    },
+    {
+      "epoch": 0.0023966924369533643,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00036920000000000003,
+      "loss": 4.3148,
+      "step": 47
+    },
+    {
+      "epoch": 0.0024476858930587548,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0003728,
+      "loss": 4.2682,
+      "step": 48
+    },
+    {
+      "epoch": 0.0024986793491641457,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00037640000000000004,
+      "loss": 4.2678,
+      "step": 49
+    },
+    {
+      "epoch": 0.002549672805269536,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00038,
+      "loss": 4.2179,
+      "step": 50
+    },
+    {
+      "epoch": 0.002600666261374927,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0003836,
+      "loss": 4.1898,
+      "step": 51
+    },
+    {
+      "epoch": 0.0026516597174803176,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0003872,
+      "loss": 4.1506,
+      "step": 52
+    },
+    {
+      "epoch": 0.0027026531735857085,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00039080000000000007,
+      "loss": 4.1246,
+      "step": 53
+    },
+    {
+      "epoch": 0.002753646629691099,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0003944,
+      "loss": 4.1309,
+      "step": 54
+    },
+    {
+      "epoch": 0.00280464008579649,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.000398,
+      "loss": 4.0911,
+      "step": 55
+    },
+    {
+      "epoch": 0.002855633541901881,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0004016,
+      "loss": 4.0603,
+      "step": 56
+    },
+    {
+      "epoch": 0.0029066269980072713,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00040520000000000004,
+      "loss": 4.0306,
+      "step": 57
+    },
+    {
+      "epoch": 0.002957620454112662,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00040880000000000007,
+      "loss": 4.0074,
+      "step": 58
+    },
+    {
+      "epoch": 0.0030086139102180527,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0004124,
+      "loss": 4.0142,
+      "step": 59
+    },
+    {
+      "epoch": 0.0030596073663234436,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00041600000000000003,
+      "loss": 3.9651,
+      "step": 60
+    },
+    {
+      "epoch": 0.003110600822428834,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0004196,
+      "loss": 3.946,
+      "step": 61
+    },
+    {
+      "epoch": 0.003161594278534225,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00042320000000000004,
+      "loss": 3.9458,
+      "step": 62
+    },
+    {
+      "epoch": 0.003212587734639616,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0004268,
+      "loss": 3.9039,
+      "step": 63
+    },
+    {
+      "epoch": 0.0032635811907450064,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0004304,
+      "loss": 3.8622,
+      "step": 64
+    },
+    {
+      "epoch": 0.0033145746468503973,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0004340000000000001,
+      "loss": 3.888,
+      "step": 65
+    },
+    {
+      "epoch": 0.0033655681029557878,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0004376,
+      "loss": 3.8407,
+      "step": 66
+    },
+    {
+      "epoch": 0.0034165615590611787,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00044120000000000004,
+      "loss": 3.8367,
+      "step": 67
+    },
+    {
+      "epoch": 0.003467555015166569,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0004448,
+      "loss": 3.8207,
+      "step": 68
+    },
+    {
+      "epoch": 0.00351854847127196,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00044840000000000006,
+      "loss": 3.8132,
+      "step": 69
+    },
+    {
+      "epoch": 0.003569541927377351,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0004520000000000001,
+      "loss": 3.7885,
+      "step": 70
+    },
+    {
+      "epoch": 0.0036205353834827415,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0004556,
+      "loss": 3.7732,
+      "step": 71
+    },
+    {
+      "epoch": 0.0036715288395881324,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0004592,
+      "loss": 3.7169,
+      "step": 72
+    },
+    {
+      "epoch": 0.003722522295693523,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0004628,
+      "loss": 3.7364,
+      "step": 73
+    },
+    {
+      "epoch": 0.0037735157517989138,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0004664,
+      "loss": 3.6857,
+      "step": 74
+    },
+    {
+      "epoch": 0.0038245092079043043,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00047000000000000004,
+      "loss": 3.6976,
+      "step": 75
+    },
+    {
+      "epoch": 0.003875502664009695,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0004736,
+      "loss": 3.6904,
+      "step": 76
+    },
+    {
+      "epoch": 0.003926496120115086,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0004772,
+      "loss": 3.6725,
+      "step": 77
+    },
+    {
+      "epoch": 0.003977489576220477,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00048080000000000003,
+      "loss": 3.6591,
+      "step": 78
+    },
+    {
+      "epoch": 0.004028483032325867,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0004844,
+      "loss": 3.6406,
+      "step": 79
+    },
+    {
+      "epoch": 0.004079476488431258,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00048800000000000004,
+      "loss": 3.632,
+      "step": 80
+    },
+    {
+      "epoch": 0.004130469944536649,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0004916,
+      "loss": 3.6319,
+      "step": 81
+    },
+    {
+      "epoch": 0.004181463400642039,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004952,
+      "loss": 3.6085,
+      "step": 82
+    },
+    {
+      "epoch": 0.00423245685674743,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004988,
+      "loss": 3.5962,
+      "step": 83
+    },
+    {
+      "epoch": 0.004283450312852821,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0005024,
+      "loss": 3.5964,
+      "step": 84
+    },
+    {
+      "epoch": 0.004334443768958212,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000506,
+      "loss": 3.565,
+      "step": 85
+    },
+    {
+      "epoch": 0.004385437225063602,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0005096,
+      "loss": 3.5736,
+      "step": 86
+    },
+    {
+      "epoch": 0.0044364306811689935,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0005132,
+      "loss": 3.561,
+      "step": 87
+    },
+    {
+      "epoch": 0.004487424137274384,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005168,
+      "loss": 3.5242,
+      "step": 88
+    },
+    {
+      "epoch": 0.0045384175933797745,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0005204,
+      "loss": 3.5322,
+      "step": 89
+    },
+    {
+      "epoch": 0.004589411049485165,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000524,
+      "loss": 3.5216,
+      "step": 90
+    },
+    {
+      "epoch": 0.004640404505590556,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0005276,
+      "loss": 3.5062,
+      "step": 91
+    },
+    {
+      "epoch": 0.004691397961695947,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0005312,
+      "loss": 3.5369,
+      "step": 92
+    },
+    {
+      "epoch": 0.004742391417801337,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0005348,
+      "loss": 3.4759,
+      "step": 93
+    },
+    {
+      "epoch": 0.004793384873906729,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0005384,
+      "loss": 3.4995,
+      "step": 94
+    },
+    {
+      "epoch": 0.004844378330012119,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0005420000000000001,
+      "loss": 3.4726,
+      "step": 95
+    },
+    {
+      "epoch": 0.0048953717861175096,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0005456000000000001,
+      "loss": 3.4882,
+      "step": 96
+    },
+    {
+      "epoch": 0.0049463652422229,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0005492,
+      "loss": 3.4805,
+      "step": 97
+    },
+    {
+      "epoch": 0.004997358698328291,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0005528,
+      "loss": 3.4831,
+      "step": 98
+    },
+    {
+      "epoch": 0.005048352154433682,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0005564,
+      "loss": 3.4418,
+      "step": 99
+    },
+    {
+      "epoch": 0.005099345610539072,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0005600000000000001,
+      "loss": 3.4178,
+      "step": 100
+    },
+    {
+      "epoch": 0.005150339066644464,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0005636000000000001,
+      "loss": 3.4542,
+      "step": 101
+    },
+    {
+      "epoch": 0.005201332522749854,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0005671999999999999,
+      "loss": 3.4302,
+      "step": 102
+    },
+    {
+      "epoch": 0.005252325978855245,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005708,
+      "loss": 3.4236,
+      "step": 103
+    },
+    {
+      "epoch": 0.005303319434960635,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0005744000000000001,
+      "loss": 3.3988,
+      "step": 104
+    },
+    {
+      "epoch": 0.0053543128910660265,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0005780000000000001,
+      "loss": 3.4117,
+      "step": 105
+    },
+    {
+      "epoch": 0.005405306347171417,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0005816,
+      "loss": 3.381,
+      "step": 106
+    },
+    {
+      "epoch": 0.0054562998032768074,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0005851999999999999,
+      "loss": 3.406,
+      "step": 107
+    },
+    {
+      "epoch": 0.005507293259382198,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0005888,
+      "loss": 3.3775,
+      "step": 108
+    },
+    {
+      "epoch": 0.005558286715487589,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005924000000000001,
+      "loss": 3.391,
+      "step": 109
+    },
+    {
+      "epoch": 0.00560928017159298,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0005960000000000001,
+      "loss": 3.3724,
+      "step": 110
+    },
+    {
+      "epoch": 0.00566027362769837,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005996,
+      "loss": 3.3516,
+      "step": 111
+    },
+    {
+      "epoch": 0.005711267083803762,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0006031999999999999,
+      "loss": 3.3698,
+      "step": 112
+    },
+    {
+      "epoch": 0.005762260539909152,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0006068,
+      "loss": 3.3634,
+      "step": 113
+    },
+    {
+      "epoch": 0.0058132539960145425,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0006104000000000001,
+      "loss": 3.3526,
+      "step": 114
+    },
+    {
+      "epoch": 0.005864247452119933,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0006140000000000001,
+      "loss": 3.3408,
+      "step": 115
+    },
+    {
+      "epoch": 0.005915240908225324,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0006176,
+      "loss": 3.3465,
+      "step": 116
+    },
+    {
+      "epoch": 0.005966234364330715,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0006212,
+      "loss": 3.328,
+      "step": 117
+    },
+    {
+      "epoch": 0.006017227820436105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0006248,
+      "loss": 3.3422,
+      "step": 118
+    },
+    {
+      "epoch": 0.006068221276541497,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0006284000000000001,
+      "loss": 3.3307,
+      "step": 119
+    },
+    {
+      "epoch": 0.006119214732646887,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.000632,
+      "loss": 3.3195,
+      "step": 120
+    },
+    {
+      "epoch": 0.006170208188752278,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0006355999999999999,
+      "loss": 3.3311,
+      "step": 121
+    },
+    {
+      "epoch": 0.006221201644857668,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0006392,
+      "loss": 3.3334,
+      "step": 122
+    },
+    {
+      "epoch": 0.0062721951009630595,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0006428,
+      "loss": 3.3016,
+      "step": 123
+    },
+    {
+      "epoch": 0.00632318855706845,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0006464000000000001,
+      "loss": 3.309,
+      "step": 124
+    },
+    {
+      "epoch": 0.00637418201317384,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 3.297,
+      "step": 125
+    },
+    {
+      "epoch": 0.006425175469279232,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0006536,
+      "loss": 3.2817,
+      "step": 126
+    },
+    {
+      "epoch": 0.006476168925384622,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0006572,
+      "loss": 3.292,
+      "step": 127
+    },
+    {
+      "epoch": 0.006527162381490013,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0006608,
+      "loss": 3.3105,
+      "step": 128
+    },
+    {
+      "epoch": 0.006578155837595403,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0006644000000000001,
+      "loss": 3.2784,
+      "step": 129
+    },
+    {
+      "epoch": 0.006629149293700795,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0006680000000000001,
+      "loss": 3.2705,
+      "step": 130
+    },
+    {
+      "epoch": 0.006680142749806185,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0006716,
+      "loss": 3.284,
+      "step": 131
+    },
+    {
+      "epoch": 0.0067311362059115755,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0006752,
+      "loss": 3.2668,
+      "step": 132
+    },
+    {
+      "epoch": 0.006782129662016967,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0006788000000000001,
+      "loss": 3.249,
+      "step": 133
+    },
+    {
+      "epoch": 0.006833123118122357,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0006824000000000001,
+      "loss": 3.2804,
+      "step": 134
+    },
+    {
+      "epoch": 0.006884116574227748,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0006860000000000001,
+      "loss": 3.2658,
+      "step": 135
+    },
+    {
+      "epoch": 0.006935110030333138,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0006896,
+      "loss": 3.2671,
+      "step": 136
+    },
+    {
+      "epoch": 0.00698610348643853,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0006932,
+      "loss": 3.2403,
+      "step": 137
+    },
+    {
+      "epoch": 0.00703709694254392,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0006968000000000001,
+      "loss": 3.2498,
+      "step": 138
+    },
+    {
+      "epoch": 0.007088090398649311,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0007004000000000001,
+      "loss": 3.2555,
+      "step": 139
+    },
+    {
+      "epoch": 0.007139083854754702,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0007040000000000002,
+      "loss": 3.2472,
+      "step": 140
+    },
+    {
+      "epoch": 0.0071900773108600925,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0007076000000000001,
+      "loss": 3.2438,
+      "step": 141
+    },
+    {
+      "epoch": 0.007241070766965483,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0007112,
+      "loss": 3.2254,
+      "step": 142
+    },
+    {
+      "epoch": 0.007292064223070873,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0007147999999999999,
+      "loss": 3.255,
+      "step": 143
+    },
+    {
+      "epoch": 0.007343057679176265,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0007183999999999999,
+      "loss": 3.2204,
+      "step": 144
+    },
+    {
+      "epoch": 0.007394051135281655,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000722,
+      "loss": 3.246,
+      "step": 145
+    },
+    {
+      "epoch": 0.007445044591387046,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0007256000000000001,
+      "loss": 3.2036,
+      "step": 146
+    },
+    {
+      "epoch": 0.007496038047492436,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0007292,
+      "loss": 3.1965,
+      "step": 147
+    },
+    {
+      "epoch": 0.0075470315035978276,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0007327999999999999,
+      "loss": 3.2063,
+      "step": 148
+    },
+    {
+      "epoch": 0.007598024959703218,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0007364,
+      "loss": 3.2127,
+      "step": 149
+    },
+    {
+      "epoch": 0.0076490184158086085,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00074,
+      "loss": 3.2028,
+      "step": 150
+    },
+    {
+      "epoch": 0.007700011871914,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0007436000000000001,
+      "loss": 3.2148,
+      "step": 151
+    },
+    {
+      "epoch": 0.00775100532801939,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0007472000000000001,
+      "loss": 3.187,
+      "step": 152
+    },
+    {
+      "epoch": 0.007801998784124781,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0007507999999999999,
+      "loss": 3.2197,
+      "step": 153
+    },
+    {
+      "epoch": 0.007852992240230172,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0007544,
+      "loss": 3.2039,
+      "step": 154
+    },
+    {
+      "epoch": 0.007903985696335563,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000758,
+      "loss": 3.1809,
+      "step": 155
+    },
+    {
+      "epoch": 0.007954979152440953,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0007616000000000001,
+      "loss": 3.2186,
+      "step": 156
+    },
+    {
+      "epoch": 0.008005972608546344,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0007652000000000002,
+      "loss": 3.201,
+      "step": 157
+    },
+    {
+      "epoch": 0.008056966064651734,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0007687999999999999,
+      "loss": 3.1765,
+      "step": 158
+    },
+    {
+      "epoch": 0.008107959520757125,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0007724,
+      "loss": 3.1859,
+      "step": 159
+    },
+    {
+      "epoch": 0.008158952976862517,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.000776,
+      "loss": 3.1707,
+      "step": 160
+    },
+    {
+      "epoch": 0.008209946432967907,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0007796000000000001,
+      "loss": 3.1949,
+      "step": 161
+    },
+    {
+      "epoch": 0.008260939889073298,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0007832000000000002,
+      "loss": 3.1744,
+      "step": 162
+    },
+    {
+      "epoch": 0.008311933345178688,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0007867999999999999,
+      "loss": 3.1757,
+      "step": 163
+    },
+    {
+      "epoch": 0.008362926801284079,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0007904,
+      "loss": 3.143,
+      "step": 164
+    },
+    {
+      "epoch": 0.00841392025738947,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0007940000000000001,
+      "loss": 3.1517,
+      "step": 165
+    },
+    {
+      "epoch": 0.00846491371349486,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0007976000000000001,
+      "loss": 3.1831,
+      "step": 166
+    },
+    {
+      "epoch": 0.008515907169600252,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0008012000000000002,
+      "loss": 3.163,
+      "step": 167
+    },
+    {
+      "epoch": 0.008566900625705642,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0008047999999999999,
+      "loss": 3.2065,
+      "step": 168
+    },
+    {
+      "epoch": 0.008617894081811033,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0008084,
+      "loss": 3.1787,
+      "step": 169
+    },
+    {
+      "epoch": 0.008668887537916423,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0008120000000000001,
+      "loss": 3.199,
+      "step": 170
+    },
+    {
+      "epoch": 0.008719880994021814,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0008156000000000001,
+      "loss": 3.1331,
+      "step": 171
+    },
+    {
+      "epoch": 0.008770874450127204,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0008192,
+      "loss": 3.1202,
+      "step": 172
+    },
+    {
+      "epoch": 0.008821867906232595,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0008227999999999999,
+      "loss": 3.1791,
+      "step": 173
+    },
+    {
+      "epoch": 0.008872861362337987,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0008264,
+      "loss": 3.1607,
+      "step": 174
+    },
+    {
+      "epoch": 0.008923854818443377,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0008300000000000001,
+      "loss": 3.1585,
+      "step": 175
+    },
+    {
+      "epoch": 0.008974848274548768,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0008335999999999999,
+      "loss": 3.1414,
+      "step": 176
+    },
+    {
+      "epoch": 0.009025841730654158,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0008372,
+      "loss": 3.1707,
+      "step": 177
+    },
+    {
+      "epoch": 0.009076835186759549,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0008408000000000001,
+      "loss": 3.1508,
+      "step": 178
+    },
+    {
+      "epoch": 0.00912782864286494,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0008444,
+      "loss": 3.1403,
+      "step": 179
+    },
+    {
+      "epoch": 0.00917882209897033,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0008480000000000001,
+      "loss": 3.1614,
+      "step": 180
+    },
+    {
+      "epoch": 0.009229815555075722,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0008515999999999999,
+      "loss": 3.1318,
+      "step": 181
+    },
+    {
+      "epoch": 0.009280809011181113,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0008552,
+      "loss": 3.135,
+      "step": 182
+    },
+    {
+      "epoch": 0.009331802467286503,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0008588000000000001,
+      "loss": 3.1438,
+      "step": 183
+    },
+    {
+      "epoch": 0.009382795923391894,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0008624,
+      "loss": 3.1199,
+      "step": 184
+    },
+    {
+      "epoch": 0.009433789379497284,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0008660000000000001,
+      "loss": 3.1227,
+      "step": 185
+    },
+    {
+      "epoch": 0.009484782835602674,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0008696,
+      "loss": 3.1425,
+      "step": 186
+    },
+    {
+      "epoch": 0.009535776291708065,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0008732,
+      "loss": 3.1292,
+      "step": 187
+    },
+    {
+      "epoch": 0.009586769747813457,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0008768000000000001,
+      "loss": 3.1429,
+      "step": 188
+    },
+    {
+      "epoch": 0.009637763203918848,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0008804,
+      "loss": 3.1073,
+      "step": 189
+    },
+    {
+      "epoch": 0.009688756660024238,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0008840000000000001,
+      "loss": 3.1163,
+      "step": 190
+    },
+    {
+      "epoch": 0.009739750116129629,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0008876,
+      "loss": 3.1352,
+      "step": 191
+    },
+    {
+      "epoch": 0.009790743572235019,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0008912,
+      "loss": 3.1119,
+      "step": 192
+    },
+    {
+      "epoch": 0.00984173702834041,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0008948000000000001,
+      "loss": 3.1143,
+      "step": 193
+    },
+    {
+      "epoch": 0.0098927304844458,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0008984000000000002,
+      "loss": 3.0917,
+      "step": 194
+    },
+    {
+      "epoch": 0.009943723940551192,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0009020000000000001,
+      "loss": 3.1264,
+      "step": 195
+    },
+    {
+      "epoch": 0.009994717396656583,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0009056,
+      "loss": 3.1428,
+      "step": 196
+    },
+    {
+      "epoch": 0.010045710852761973,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0009092,
+      "loss": 3.1288,
+      "step": 197
+    },
+    {
+      "epoch": 0.010096704308867364,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0009128000000000001,
+      "loss": 3.1118,
+      "step": 198
+    },
+    {
+      "epoch": 0.010147697764972754,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0009164000000000002,
+      "loss": 3.1073,
+      "step": 199
+    },
+    {
+      "epoch": 0.010198691221078145,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0009200000000000001,
+      "loss": 3.1353,
+      "step": 200
+    },
+    {
+      "epoch": 0.010249684677183535,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0009236,
+      "loss": 3.1067,
+      "step": 201
+    },
+    {
+      "epoch": 0.010300678133288927,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0009272000000000001,
+      "loss": 3.0886,
+      "step": 202
+    },
+    {
+      "epoch": 0.010351671589394318,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0009308000000000001,
+      "loss": 3.1173,
+      "step": 203
+    },
+    {
+      "epoch": 0.010402665045499708,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0009343999999999999,
+      "loss": 3.0982,
+      "step": 204
+    },
+    {
+      "epoch": 0.010453658501605099,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0009379999999999999,
+      "loss": 3.1175,
+      "step": 205
+    },
+    {
+      "epoch": 0.01050465195771049,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0009416,
+      "loss": 3.0875,
+      "step": 206
+    },
+    {
+      "epoch": 0.01055564541381588,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0009452000000000001,
+      "loss": 3.0999,
+      "step": 207
+    },
+    {
+      "epoch": 0.01060663886992127,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0009488000000000001,
+      "loss": 3.0867,
+      "step": 208
+    },
+    {
+      "epoch": 0.01065763232602666,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0009524,
+      "loss": 3.0711,
+      "step": 209
+    },
+    {
+      "epoch": 0.010708625782132053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0009559999999999999,
+      "loss": 3.1125,
+      "step": 210
+    },
+    {
+      "epoch": 0.010759619238237443,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0009596,
+      "loss": 3.0881,
+      "step": 211
+    },
+    {
+      "epoch": 0.010810612694342834,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0009632000000000001,
+      "loss": 3.1159,
+      "step": 212
+    },
+    {
+      "epoch": 0.010861606150448224,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0009668000000000001,
+      "loss": 3.0959,
+      "step": 213
+    },
+    {
+      "epoch": 0.010912599606553615,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0009704,
+      "loss": 3.1047,
+      "step": 214
+    },
+    {
+      "epoch": 0.010963593062659005,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000974,
+      "loss": 3.0925,
+      "step": 215
+    },
+    {
+      "epoch": 0.011014586518764396,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0009776000000000001,
+      "loss": 3.0962,
+      "step": 216
+    },
+    {
+      "epoch": 0.011065579974869788,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0009812,
+      "loss": 3.093,
+      "step": 217
+    },
+    {
+      "epoch": 0.011116573430975179,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0009848,
+      "loss": 3.0794,
+      "step": 218
+    },
+    {
+      "epoch": 0.011167566887080569,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0009884,
+      "loss": 3.0923,
+      "step": 219
+    },
+    {
+      "epoch": 0.01121856034318596,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000992,
+      "loss": 3.081,
+      "step": 220
+    },
+    {
+      "epoch": 0.01126955379929135,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0009956000000000001,
+      "loss": 3.0678,
+      "step": 221
+    },
+    {
+      "epoch": 0.01132054725539674,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0009992,
+      "loss": 3.0905,
+      "step": 222
+    },
+    {
+      "epoch": 0.011371540711502131,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0010028,
+      "loss": 3.0869,
+      "step": 223
+    },
+    {
+      "epoch": 0.011422534167607523,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0010064,
+      "loss": 3.0967,
+      "step": 224
+    },
+    {
+      "epoch": 0.011473527623712914,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00101,
+      "loss": 3.0911,
+      "step": 225
+    },
+    {
+      "epoch": 0.011524521079818304,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0010136000000000001,
+      "loss": 3.0759,
+      "step": 226
+    },
+    {
+      "epoch": 0.011575514535923695,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0010172000000000002,
+      "loss": 3.093,
+      "step": 227
+    },
+    {
+      "epoch": 0.011626507992029085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0010208,
+      "loss": 3.112,
+      "step": 228
+    },
+    {
+      "epoch": 0.011677501448134476,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0010244,
+      "loss": 3.0587,
+      "step": 229
+    },
+    {
+      "epoch": 0.011728494904239866,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.001028,
+      "loss": 3.0898,
+      "step": 230
+    },
+    {
+      "epoch": 0.011779488360345258,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0010316000000000001,
+      "loss": 3.0754,
+      "step": 231
+    },
+    {
+      "epoch": 0.011830481816450649,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0010352000000000002,
+      "loss": 3.0914,
+      "step": 232
+    },
+    {
+      "epoch": 0.01188147527255604,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0010388,
+      "loss": 3.1138,
+      "step": 233
+    },
+    {
+      "epoch": 0.01193246872866143,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0010424,
+      "loss": 3.0728,
+      "step": 234
+    },
+    {
+      "epoch": 0.01198346218476682,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.001046,
+      "loss": 3.0821,
+      "step": 235
+    },
+    {
+      "epoch": 0.01203445564087221,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0010496000000000001,
+      "loss": 3.0796,
+      "step": 236
+    },
+    {
+      "epoch": 0.012085449096977601,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0010532,
+      "loss": 3.0582,
+      "step": 237
+    },
+    {
+      "epoch": 0.012136442553082993,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0010567999999999999,
+      "loss": 3.0566,
+      "step": 238
+    },
+    {
+      "epoch": 0.012187436009188384,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0010604,
+      "loss": 3.1031,
+      "step": 239
+    },
+    {
+      "epoch": 0.012238429465293774,
+      "grad_norm": 0.5,
+      "learning_rate": 0.001064,
+      "loss": 3.0619,
+      "step": 240
+    },
+    {
+      "epoch": 0.012289422921399165,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0010676000000000001,
+      "loss": 3.0683,
+      "step": 241
+    },
+    {
+      "epoch": 0.012340416377504555,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0010712,
+      "loss": 3.0639,
+      "step": 242
+    },
+    {
+      "epoch": 0.012391409833609946,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0010748,
+      "loss": 3.0523,
+      "step": 243
+    },
+    {
+      "epoch": 0.012442403289715336,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0010784,
+      "loss": 3.0639,
+      "step": 244
+    },
+    {
+      "epoch": 0.012493396745820728,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.001082,
+      "loss": 3.0765,
+      "step": 245
+    },
+    {
+      "epoch": 0.012544390201926119,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0010856000000000001,
+      "loss": 3.0657,
+      "step": 246
+    },
+    {
+      "epoch": 0.01259538365803151,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0010892,
+      "loss": 3.077,
+      "step": 247
+    },
+    {
+      "epoch": 0.0126463771141369,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0010928,
+      "loss": 3.0561,
+      "step": 248
+    },
+    {
+      "epoch": 0.01269737057024229,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0010964,
+      "loss": 3.0575,
+      "step": 249
+    },
+    {
+      "epoch": 0.01274836402634768,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0011,
+      "loss": 3.0866,
+      "step": 250
+    },
+    {
+      "epoch": 0.012799357482453071,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0011036000000000002,
+      "loss": 3.0746,
+      "step": 251
+    },
+    {
+      "epoch": 0.012850350938558464,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0011072,
+      "loss": 3.0545,
+      "step": 252
+    },
+    {
+      "epoch": 0.012901344394663854,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0011108000000000001,
+      "loss": 3.0542,
+      "step": 253
+    },
+    {
+      "epoch": 0.012952337850769245,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0011144,
+      "loss": 3.0648,
+      "step": 254
+    },
+    {
+      "epoch": 0.013003331306874635,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.001118,
+      "loss": 3.0665,
+      "step": 255
+    },
+    {
+      "epoch": 0.013054324762980025,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0011216000000000002,
+      "loss": 3.0515,
+      "step": 256
+    },
+    {
+      "epoch": 0.013105318219085416,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0011252,
+      "loss": 3.0518,
+      "step": 257
+    },
+    {
+      "epoch": 0.013156311675190806,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0011288000000000001,
+      "loss": 3.0824,
+      "step": 258
+    },
+    {
+      "epoch": 0.013207305131296199,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0011324000000000002,
+      "loss": 3.0969,
+      "step": 259
+    },
+    {
+      "epoch": 0.01325829858740159,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.001136,
+      "loss": 3.0713,
+      "step": 260
+    },
+    {
+      "epoch": 0.01330929204350698,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0011396000000000002,
+      "loss": 3.0805,
+      "step": 261
+    },
+    {
+      "epoch": 0.01336028549961237,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0011432,
+      "loss": 3.0451,
+      "step": 262
+    },
+    {
+      "epoch": 0.01341127895571776,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0011468000000000001,
+      "loss": 3.0859,
+      "step": 263
+    },
+    {
+      "epoch": 0.013462272411823151,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0011504000000000002,
+      "loss": 3.0711,
+      "step": 264
+    },
+    {
+      "epoch": 0.013513265867928542,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.001154,
+      "loss": 3.0798,
+      "step": 265
+    },
+    {
+      "epoch": 0.013564259324033934,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0011576000000000002,
+      "loss": 3.0662,
+      "step": 266
+    },
+    {
+      "epoch": 0.013615252780139324,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0011612,
+      "loss": 3.0997,
+      "step": 267
+    },
+    {
+      "epoch": 0.013666246236244715,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0011648000000000001,
+      "loss": 3.0509,
+      "step": 268
+    },
+    {
+      "epoch": 0.013717239692350105,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0011684000000000002,
+      "loss": 3.0507,
+      "step": 269
+    },
+    {
+      "epoch": 0.013768233148455496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.001172,
+      "loss": 3.0811,
+      "step": 270
+    },
+    {
+      "epoch": 0.013819226604560886,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0011756000000000002,
+      "loss": 3.046,
+      "step": 271
+    },
+    {
+      "epoch": 0.013870220060666277,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0011792,
+      "loss": 3.0685,
+      "step": 272
+    },
+    {
+      "epoch": 0.013921213516771669,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0011828000000000001,
+      "loss": 3.0454,
+      "step": 273
+    },
+    {
+      "epoch": 0.01397220697287706,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0011864000000000002,
+      "loss": 3.0524,
+      "step": 274
+    },
+    {
+      "epoch": 0.01402320042898245,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00119,
+      "loss": 3.0407,
+      "step": 275
+    },
+    {
+      "epoch": 0.01407419388508784,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0011936000000000002,
+      "loss": 3.0317,
+      "step": 276
+    },
+    {
+      "epoch": 0.01412518734119323,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0011972,
+      "loss": 3.0485,
+      "step": 277
+    },
+    {
+      "epoch": 0.014176180797298621,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0012008000000000001,
+      "loss": 3.0345,
+      "step": 278
+    },
+    {
+      "epoch": 0.014227174253404012,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0012044000000000002,
+      "loss": 3.0852,
+      "step": 279
+    },
+    {
+      "epoch": 0.014278167709509404,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0012080000000000003,
+      "loss": 3.0495,
+      "step": 280
+    },
+    {
+      "epoch": 0.014329161165614794,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0012116,
+      "loss": 3.0537,
+      "step": 281
+    },
+    {
+      "epoch": 0.014380154621720185,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0012151999999999998,
+      "loss": 3.0525,
+      "step": 282
+    },
+    {
+      "epoch": 0.014431148077825575,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0012188,
+      "loss": 3.0518,
+      "step": 283
+    },
+    {
+      "epoch": 0.014482141533930966,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0012224,
+      "loss": 3.0568,
+      "step": 284
+    },
+    {
+      "epoch": 0.014533134990036356,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.001226,
+      "loss": 3.0595,
+      "step": 285
+    },
+    {
+      "epoch": 0.014584128446141747,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0012295999999999997,
+      "loss": 3.0558,
+      "step": 286
+    },
+    {
+      "epoch": 0.014635121902247139,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0012331999999999998,
+      "loss": 3.0187,
+      "step": 287
+    },
+    {
+      "epoch": 0.01468611535835253,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0012368,
+      "loss": 3.0654,
+      "step": 288
+    },
+    {
+      "epoch": 0.01473710881445792,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0012404,
+      "loss": 3.0127,
+      "step": 289
+    },
+    {
+      "epoch": 0.01478810227056331,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.001244,
+      "loss": 3.0508,
+      "step": 290
+    },
+    {
+      "epoch": 0.014839095726668701,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0012475999999999998,
+      "loss": 3.0767,
+      "step": 291
+    },
+    {
+      "epoch": 0.014890089182774091,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0012511999999999998,
+      "loss": 3.0618,
+      "step": 292
+    },
+    {
+      "epoch": 0.014941082638879482,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0012548,
+      "loss": 3.0458,
+      "step": 293
+    },
+    {
+      "epoch": 0.014992076094984872,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0012584,
+      "loss": 3.0447,
+      "step": 294
+    },
+    {
+      "epoch": 0.015043069551090265,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.001262,
+      "loss": 3.0232,
+      "step": 295
+    },
+    {
+      "epoch": 0.015094063007195655,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0012655999999999998,
+      "loss": 3.035,
+      "step": 296
+    },
+    {
+      "epoch": 0.015145056463301046,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0012691999999999998,
+      "loss": 3.035,
+      "step": 297
+    },
+    {
+      "epoch": 0.015196049919406436,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0012728,
+      "loss": 3.0354,
+      "step": 298
+    },
+    {
+      "epoch": 0.015247043375511827,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0012764,
+      "loss": 3.0447,
+      "step": 299
+    },
+    {
+      "epoch": 0.015298036831617217,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00128,
+      "loss": 3.0468,
+      "step": 300
+    },
+    {
+      "epoch": 0.015349030287722608,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0012836,
+      "loss": 3.0292,
+      "step": 301
+    },
+    {
+      "epoch": 0.015400023743828,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0012871999999999998,
+      "loss": 3.0251,
+      "step": 302
+    },
+    {
+      "epoch": 0.01545101719993339,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0012908,
+      "loss": 3.0617,
+      "step": 303
+    },
+    {
+      "epoch": 0.01550201065603878,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0012944,
+      "loss": 3.0488,
+      "step": 304
+    },
+    {
+      "epoch": 0.015553004112144171,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0012980000000000001,
+      "loss": 3.0642,
+      "step": 305
+    },
+    {
+      "epoch": 0.015603997568249562,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0013016,
+      "loss": 3.045,
+      "step": 306
+    },
+    {
+      "epoch": 0.015654991024354952,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0013051999999999998,
+      "loss": 3.0433,
+      "step": 307
+    },
+    {
+      "epoch": 0.015705984480460344,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0013088,
+      "loss": 3.0607,
+      "step": 308
+    },
+    {
+      "epoch": 0.015756977936565733,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0013124,
+      "loss": 3.0536,
+      "step": 309
+    },
+    {
+      "epoch": 0.015807971392671125,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0013160000000000001,
+      "loss": 3.0269,
+      "step": 310
+    },
+    {
+      "epoch": 0.015858964848776514,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0013196,
+      "loss": 3.0281,
+      "step": 311
+    },
+    {
+      "epoch": 0.015909958304881906,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0013231999999999999,
+      "loss": 3.0137,
+      "step": 312
+    },
+    {
+      "epoch": 0.0159609517609873,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0013268,
+      "loss": 3.0423,
+      "step": 313
+    },
+    {
+      "epoch": 0.016011945217092687,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0013304,
+      "loss": 3.0389,
+      "step": 314
+    },
+    {
+      "epoch": 0.01606293867319808,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0013340000000000001,
+      "loss": 3.0255,
+      "step": 315
+    },
+    {
+      "epoch": 0.016113932129303468,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0013376,
+      "loss": 3.0225,
+      "step": 316
+    },
+    {
+      "epoch": 0.01616492558540886,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0013412,
+      "loss": 3.0203,
+      "step": 317
+    },
+    {
+      "epoch": 0.01621591904151425,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0013448,
+      "loss": 3.041,
+      "step": 318
+    },
+    {
+      "epoch": 0.01626691249761964,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0013484,
+      "loss": 3.0099,
+      "step": 319
+    },
+    {
+      "epoch": 0.016317905953725034,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0013520000000000001,
+      "loss": 3.0549,
+      "step": 320
+    },
+    {
+      "epoch": 0.016368899409830422,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0013556,
+      "loss": 3.0404,
+      "step": 321
+    },
+    {
+      "epoch": 0.016419892865935815,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0013592,
+      "loss": 3.0351,
+      "step": 322
+    },
+    {
+      "epoch": 0.016470886322041203,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0013628,
+      "loss": 3.0374,
+      "step": 323
+    },
+    {
+      "epoch": 0.016521879778146596,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0013664,
+      "loss": 3.0454,
+      "step": 324
+    },
+    {
+      "epoch": 0.016572873234251984,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0013700000000000001,
+      "loss": 3.0545,
+      "step": 325
+    },
+    {
+      "epoch": 0.016623866690357376,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0013736,
+      "loss": 3.0536,
+      "step": 326
+    },
+    {
+      "epoch": 0.01667486014646277,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0013772,
+      "loss": 3.0029,
+      "step": 327
+    },
+    {
+      "epoch": 0.016725853602568157,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0013808,
+      "loss": 3.0196,
+      "step": 328
+    },
+    {
+      "epoch": 0.01677684705867355,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0013844,
+      "loss": 3.0125,
+      "step": 329
+    },
+    {
+      "epoch": 0.01682784051477894,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0013880000000000001,
+      "loss": 3.0403,
+      "step": 330
+    },
+    {
+      "epoch": 0.01687883397088433,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0013916,
+      "loss": 3.0608,
+      "step": 331
+    },
+    {
+      "epoch": 0.01692982742698972,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0013952,
+      "loss": 3.0226,
+      "step": 332
+    },
+    {
+      "epoch": 0.01698082088309511,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0013988000000000002,
+      "loss": 3.0434,
+      "step": 333
+    },
+    {
+      "epoch": 0.017031814339200504,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0014024,
+      "loss": 3.0189,
+      "step": 334
+    },
+    {
+      "epoch": 0.017082807795305893,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0014060000000000001,
+      "loss": 3.051,
+      "step": 335
+    },
+    {
+      "epoch": 0.017133801251411285,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0014096,
+      "loss": 3.0133,
+      "step": 336
+    },
+    {
+      "epoch": 0.017184794707516673,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0014132,
+      "loss": 3.0276,
+      "step": 337
+    },
+    {
+      "epoch": 0.017235788163622066,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0014168000000000002,
+      "loss": 2.9947,
+      "step": 338
+    },
+    {
+      "epoch": 0.017286781619727454,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0014204,
+      "loss": 3.0234,
+      "step": 339
+    },
+    {
+      "epoch": 0.017337775075832847,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0014240000000000001,
+      "loss": 3.0296,
+      "step": 340
+    },
+    {
+      "epoch": 0.01738876853193824,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0014276,
+      "loss": 2.9919,
+      "step": 341
+    },
+    {
+      "epoch": 0.017439761988043628,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0014312,
+      "loss": 3.0203,
+      "step": 342
+    },
+    {
+      "epoch": 0.01749075544414902,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0014348000000000002,
+      "loss": 3.0362,
+      "step": 343
+    },
+    {
+      "epoch": 0.01754174890025441,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0014383999999999998,
+      "loss": 3.0311,
+      "step": 344
+    },
+    {
+      "epoch": 0.0175927423563598,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.001442,
+      "loss": 3.0297,
+      "step": 345
+    },
+    {
+      "epoch": 0.01764373581246519,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0014456,
+      "loss": 3.042,
+      "step": 346
+    },
+    {
+      "epoch": 0.017694729268570582,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0014491999999999999,
+      "loss": 3.0325,
+      "step": 347
+    },
+    {
+      "epoch": 0.017745722724675974,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0014528,
+      "loss": 3.0081,
+      "step": 348
+    },
+    {
+      "epoch": 0.017796716180781363,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0014563999999999998,
+      "loss": 3.024,
+      "step": 349
+    },
+    {
+      "epoch": 0.017847709636886755,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00146,
+      "loss": 3.0429,
+      "step": 350
+    },
+    {
+      "epoch": 0.017898703092992144,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0014636,
+      "loss": 3.0357,
+      "step": 351
+    },
+    {
+      "epoch": 0.017949696549097536,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0014671999999999999,
+      "loss": 3.0286,
+      "step": 352
+    },
+    {
+      "epoch": 0.018000690005202925,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0014708,
+      "loss": 3.0157,
+      "step": 353
+    },
+    {
+      "epoch": 0.018051683461308317,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0014744,
+      "loss": 3.0356,
+      "step": 354
+    },
+    {
+      "epoch": 0.01810267691741371,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.001478,
+      "loss": 3.0383,
+      "step": 355
+    },
+    {
+      "epoch": 0.018153670373519098,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0014816,
+      "loss": 3.0152,
+      "step": 356
+    },
+    {
+      "epoch": 0.01820466382962449,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0014851999999999999,
+      "loss": 3.0419,
+      "step": 357
+    },
+    {
+      "epoch": 0.01825565728572988,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0014888,
+      "loss": 3.0228,
+      "step": 358
+    },
+    {
+      "epoch": 0.01830665074183527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0014924,
+      "loss": 3.0077,
+      "step": 359
+    },
+    {
+      "epoch": 0.01835764419794066,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.001496,
+      "loss": 3.0024,
+      "step": 360
+    },
+    {
+      "epoch": 0.018408637654046052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0014996,
+      "loss": 3.0068,
+      "step": 361
+    },
+    {
+      "epoch": 0.018459631110151444,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0015032,
+      "loss": 3.0179,
+      "step": 362
+    },
+    {
+      "epoch": 0.018510624566256833,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0015068,
+      "loss": 3.0262,
+      "step": 363
+    },
+    {
+      "epoch": 0.018561618022362225,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0015104,
+      "loss": 3.0142,
+      "step": 364
+    },
+    {
+      "epoch": 0.018612611478467614,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001514,
+      "loss": 3.0228,
+      "step": 365
+    },
+    {
+      "epoch": 0.018663604934573006,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0015176,
+      "loss": 3.0232,
+      "step": 366
+    },
+    {
+      "epoch": 0.018714598390678395,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0015212,
+      "loss": 3.0156,
+      "step": 367
+    },
+    {
+      "epoch": 0.018765591846783787,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0015248,
+      "loss": 3.0221,
+      "step": 368
+    },
+    {
+      "epoch": 0.01881658530288918,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0015284,
+      "loss": 3.0389,
+      "step": 369
+    },
+    {
+      "epoch": 0.018867578758994568,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0015320000000000002,
+      "loss": 3.0147,
+      "step": 370
+    },
+    {
+      "epoch": 0.01891857221509996,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0015356,
+      "loss": 3.0368,
+      "step": 371
+    },
+    {
+      "epoch": 0.01896956567120535,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0015392,
+      "loss": 3.0264,
+      "step": 372
+    },
+    {
+      "epoch": 0.01902055912731074,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0015428,
+      "loss": 3.0467,
+      "step": 373
+    },
+    {
+      "epoch": 0.01907155258341613,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0015464,
+      "loss": 3.0018,
+      "step": 374
+    },
+    {
+      "epoch": 0.019122546039521522,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0015500000000000002,
+      "loss": 3.0433,
+      "step": 375
+    },
+    {
+      "epoch": 0.019173539495626914,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0015536,
+      "loss": 3.0105,
+      "step": 376
+    },
+    {
+      "epoch": 0.019224532951732303,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0015572,
+      "loss": 3.0387,
+      "step": 377
+    },
+    {
+      "epoch": 0.019275526407837695,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0015608,
+      "loss": 3.0034,
+      "step": 378
+    },
+    {
+      "epoch": 0.019326519863943084,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0015644,
+      "loss": 3.0103,
+      "step": 379
+    },
+    {
+      "epoch": 0.019377513320048476,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0015680000000000002,
+      "loss": 3.0475,
+      "step": 380
+    },
+    {
+      "epoch": 0.019428506776153865,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0015716,
+      "loss": 3.0431,
+      "step": 381
+    },
+    {
+      "epoch": 0.019479500232259257,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0015752,
+      "loss": 3.0123,
+      "step": 382
+    },
+    {
+      "epoch": 0.01953049368836465,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0015788,
+      "loss": 3.0232,
+      "step": 383
+    },
+    {
+      "epoch": 0.019581487144470038,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0015824,
+      "loss": 3.0239,
+      "step": 384
+    },
+    {
+      "epoch": 0.01963248060057543,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0015860000000000002,
+      "loss": 3.0307,
+      "step": 385
+    },
+    {
+      "epoch": 0.01968347405668082,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0015896000000000003,
+      "loss": 3.0144,
+      "step": 386
+    },
+    {
+      "epoch": 0.01973446751278621,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0015932,
+      "loss": 3.0147,
+      "step": 387
+    },
+    {
+      "epoch": 0.0197854609688916,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0015968,
+      "loss": 3.0496,
+      "step": 388
+    },
+    {
+      "epoch": 0.019836454424996992,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0016004,
+      "loss": 3.0393,
+      "step": 389
+    },
+    {
+      "epoch": 0.019887447881102385,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0016040000000000002,
+      "loss": 3.0296,
+      "step": 390
+    },
+    {
+      "epoch": 0.019938441337207773,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0016076000000000003,
+      "loss": 3.0364,
+      "step": 391
+    },
+    {
+      "epoch": 0.019989434793313166,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0016112,
+      "loss": 3.022,
+      "step": 392
+    },
+    {
+      "epoch": 0.020040428249418554,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0016148,
+      "loss": 3.0122,
+      "step": 393
+    },
+    {
+      "epoch": 0.020091421705523946,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0016184,
+      "loss": 3.0346,
+      "step": 394
+    },
+    {
+      "epoch": 0.020142415161629335,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0016220000000000002,
+      "loss": 3.0102,
+      "step": 395
+    },
+    {
+      "epoch": 0.020193408617734727,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0016256000000000003,
+      "loss": 3.0234,
+      "step": 396
+    },
+    {
+      "epoch": 0.02024440207384012,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0016292,
+      "loss": 3.0065,
+      "step": 397
+    },
+    {
+      "epoch": 0.02029539552994551,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0016328,
+      "loss": 3.0347,
+      "step": 398
+    },
+    {
+      "epoch": 0.0203463889860509,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0016364,
+      "loss": 3.0368,
+      "step": 399
+    },
+    {
+      "epoch": 0.02039738244215629,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0016400000000000002,
+      "loss": 3.0262,
+      "step": 400
+    },
+    {
+      "epoch": 0.02044837589826168,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0016436000000000003,
+      "loss": 3.0061,
+      "step": 401
+    },
+    {
+      "epoch": 0.02049936935436707,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0016472000000000001,
+      "loss": 3.0346,
+      "step": 402
+    },
+    {
+      "epoch": 0.020550362810472463,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0016508,
+      "loss": 2.9946,
+      "step": 403
+    },
+    {
+      "epoch": 0.020601356266577855,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0016544,
+      "loss": 3.0229,
+      "step": 404
+    },
+    {
+      "epoch": 0.020652349722683244,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0016580000000000002,
+      "loss": 3.0198,
+      "step": 405
+    },
+    {
+      "epoch": 0.020703343178788636,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0016616000000000003,
+      "loss": 3.042,
+      "step": 406
+    },
+    {
+      "epoch": 0.020754336634894024,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0016652,
+      "loss": 3.0228,
+      "step": 407
+    },
+    {
+      "epoch": 0.020805330090999417,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0016687999999999998,
+      "loss": 3.0175,
+      "step": 408
+    },
+    {
+      "epoch": 0.020856323547104805,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0016723999999999999,
+      "loss": 3.0293,
+      "step": 409
+    },
+    {
+      "epoch": 0.020907317003210198,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.001676,
+      "loss": 3.0114,
+      "step": 410
+    },
+    {
+      "epoch": 0.02095831045931559,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0016796,
+      "loss": 3.0236,
+      "step": 411
+    },
+    {
+      "epoch": 0.02100930391542098,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0016832,
+      "loss": 3.047,
+      "step": 412
+    },
+    {
+      "epoch": 0.02106029737152637,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0016867999999999998,
+      "loss": 3.01,
+      "step": 413
+    },
+    {
+      "epoch": 0.02111129082763176,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0016903999999999999,
+      "loss": 3.025,
+      "step": 414
+    },
+    {
+      "epoch": 0.021162284283737152,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.001694,
+      "loss": 3.0176,
+      "step": 415
+    },
+    {
+      "epoch": 0.02121327773984254,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0016976,
+      "loss": 3.0333,
+      "step": 416
+    },
+    {
+      "epoch": 0.021264271195947933,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0017012,
+      "loss": 3.0025,
+      "step": 417
+    },
+    {
+      "epoch": 0.02131526465205332,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0017047999999999998,
+      "loss": 3.0521,
+      "step": 418
+    },
+    {
+      "epoch": 0.021366258108158714,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0017084,
+      "loss": 3.0112,
+      "step": 419
+    },
+    {
+      "epoch": 0.021417251564264106,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.001712,
+      "loss": 2.9908,
+      "step": 420
+    },
+    {
+      "epoch": 0.021468245020369495,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0017156,
+      "loss": 3.0242,
+      "step": 421
+    },
+    {
+      "epoch": 0.021519238476474887,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0017192,
+      "loss": 3.0103,
+      "step": 422
+    },
+    {
+      "epoch": 0.021570231932580276,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0017228,
+      "loss": 3.0227,
+      "step": 423
+    },
+    {
+      "epoch": 0.021621225388685668,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0017264,
+      "loss": 3.0506,
+      "step": 424
+    },
+    {
+      "epoch": 0.021672218844791057,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00173,
+      "loss": 3.0198,
+      "step": 425
+    },
+    {
+      "epoch": 0.02172321230089645,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0017336,
+      "loss": 3.0352,
+      "step": 426
+    },
+    {
+      "epoch": 0.02177420575700184,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0017372,
+      "loss": 3.0257,
+      "step": 427
+    },
+    {
+      "epoch": 0.02182519921310723,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0017408,
+      "loss": 3.0278,
+      "step": 428
+    },
+    {
+      "epoch": 0.021876192669212622,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0017444,
+      "loss": 3.0035,
+      "step": 429
+    },
+    {
+      "epoch": 0.02192718612531801,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.001748,
+      "loss": 3.0124,
+      "step": 430
+    },
+    {
+      "epoch": 0.021978179581423403,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0017516,
+      "loss": 2.9945,
+      "step": 431
+    },
+    {
+      "epoch": 0.02202917303752879,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0017552,
+      "loss": 2.9998,
+      "step": 432
+    },
+    {
+      "epoch": 0.022080166493634184,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0017588,
+      "loss": 3.0111,
+      "step": 433
+    },
+    {
+      "epoch": 0.022131159949739576,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0017624,
+      "loss": 3.055,
+      "step": 434
+    },
+    {
+      "epoch": 0.022182153405844965,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001766,
+      "loss": 3.0217,
+      "step": 435
+    },
+    {
+      "epoch": 0.022233146861950357,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0017696,
+      "loss": 3.0121,
+      "step": 436
+    },
+    {
+      "epoch": 0.022284140318055746,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0017732,
+      "loss": 3.0349,
+      "step": 437
+    },
+    {
+      "epoch": 0.022335133774161138,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0017768,
+      "loss": 3.0447,
+      "step": 438
+    },
+    {
+      "epoch": 0.022386127230266527,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0017804000000000001,
+      "loss": 3.0236,
+      "step": 439
+    },
+    {
+      "epoch": 0.02243712068637192,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001784,
+      "loss": 3.0114,
+      "step": 440
+    },
+    {
+      "epoch": 0.02248811414247731,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0017876,
+      "loss": 3.0223,
+      "step": 441
+    },
+    {
+      "epoch": 0.0225391075985827,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0017912,
+      "loss": 2.9907,
+      "step": 442
+    },
+    {
+      "epoch": 0.022590101054688092,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0017948,
+      "loss": 3.0362,
+      "step": 443
+    },
+    {
+      "epoch": 0.02264109451079348,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0017984000000000001,
+      "loss": 3.001,
+      "step": 444
+    },
+    {
+      "epoch": 0.022692087966898873,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.001802,
+      "loss": 2.9964,
+      "step": 445
+    },
+    {
+      "epoch": 0.022743081423004262,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0018056,
+      "loss": 3.02,
+      "step": 446
+    },
+    {
+      "epoch": 0.022794074879109654,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0018092,
+      "loss": 3.0273,
+      "step": 447
+    },
+    {
+      "epoch": 0.022845068335215046,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0018128,
+      "loss": 3.0392,
+      "step": 448
+    },
+    {
+      "epoch": 0.022896061791320435,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0018164000000000001,
+      "loss": 3.0031,
+      "step": 449
+    },
+    {
+      "epoch": 0.022947055247425827,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00182,
+      "loss": 3.0329,
+      "step": 450
+    },
+    {
+      "epoch": 0.022998048703531216,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0018236,
+      "loss": 3.0301,
+      "step": 451
+    },
+    {
+      "epoch": 0.023049042159636608,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0018272,
+      "loss": 3.0365,
+      "step": 452
+    },
+    {
+      "epoch": 0.023100035615741997,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0018308,
+      "loss": 3.0246,
+      "step": 453
+    },
+    {
+      "epoch": 0.02315102907184739,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0018344000000000001,
+      "loss": 3.0206,
+      "step": 454
+    },
+    {
+      "epoch": 0.02320202252795278,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0018380000000000002,
+      "loss": 3.0126,
+      "step": 455
+    },
+    {
+      "epoch": 0.02325301598405817,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0018416,
+      "loss": 3.0301,
+      "step": 456
+    },
+    {
+      "epoch": 0.023304009440163562,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0018452,
+      "loss": 3.0373,
+      "step": 457
+    },
+    {
+      "epoch": 0.02335500289626895,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0018488,
+      "loss": 2.9892,
+      "step": 458
+    },
+    {
+      "epoch": 0.023405996352374343,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0018524000000000001,
+      "loss": 3.0463,
+      "step": 459
+    },
+    {
+      "epoch": 0.023456989808479732,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0018560000000000002,
+      "loss": 3.0193,
+      "step": 460
+    },
+    {
+      "epoch": 0.023507983264585124,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0018596,
+      "loss": 3.0095,
+      "step": 461
+    },
+    {
+      "epoch": 0.023558976720690517,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0018632,
+      "loss": 3.0271,
+      "step": 462
+    },
+    {
+      "epoch": 0.023609970176795905,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0018668,
+      "loss": 3.014,
+      "step": 463
+    },
+    {
+      "epoch": 0.023660963632901297,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0018704000000000001,
+      "loss": 3.0174,
+      "step": 464
+    },
+    {
+      "epoch": 0.023711957089006686,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0018740000000000002,
+      "loss": 3.0069,
+      "step": 465
+    },
+    {
+      "epoch": 0.02376295054511208,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0018776,
+      "loss": 3.0139,
+      "step": 466
+    },
+    {
+      "epoch": 0.023813944001217467,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0018812,
+      "loss": 2.9907,
+      "step": 467
+    },
+    {
+      "epoch": 0.02386493745732286,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0018848,
+      "loss": 3.0252,
+      "step": 468
+    },
+    {
+      "epoch": 0.02391593091342825,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0018884,
+      "loss": 3.0231,
+      "step": 469
+    },
+    {
+      "epoch": 0.02396692436953364,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.001892,
+      "loss": 3.004,
+      "step": 470
+    },
+    {
+      "epoch": 0.024017917825639033,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0018955999999999999,
+      "loss": 3.0186,
+      "step": 471
+    },
+    {
+      "epoch": 0.02406891128174442,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0018992,
+      "loss": 3.0181,
+      "step": 472
+    },
+    {
+      "epoch": 0.024119904737849814,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0019027999999999998,
+      "loss": 3.0253,
+      "step": 473
+    },
+    {
+      "epoch": 0.024170898193955202,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019064,
+      "loss": 2.9968,
+      "step": 474
+    },
+    {
+      "epoch": 0.024221891650060594,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00191,
+      "loss": 3.0152,
+      "step": 475
+    },
+    {
+      "epoch": 0.024272885106165987,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019136000000000001,
+      "loss": 3.0091,
+      "step": 476
+    },
+    {
+      "epoch": 0.024323878562271375,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0019172,
+      "loss": 3.0284,
+      "step": 477
+    },
+    {
+      "epoch": 0.024374872018376768,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0019207999999999999,
+      "loss": 3.0213,
+      "step": 478
+    },
+    {
+      "epoch": 0.024425865474482156,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019244,
+      "loss": 3.0326,
+      "step": 479
+    },
+    {
+      "epoch": 0.02447685893058755,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.001928,
+      "loss": 3.0193,
+      "step": 480
+    },
+    {
+      "epoch": 0.024527852386692937,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019316000000000001,
+      "loss": 3.0255,
+      "step": 481
+    },
+    {
+      "epoch": 0.02457884584279833,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019352,
+      "loss": 3.0065,
+      "step": 482
+    },
+    {
+      "epoch": 0.024629839298903722,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019387999999999999,
+      "loss": 3.0122,
+      "step": 483
+    },
+    {
+      "epoch": 0.02468083275500911,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019424,
+      "loss": 3.0257,
+      "step": 484
+    },
+    {
+      "epoch": 0.024731826211114503,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.001946,
+      "loss": 3.0215,
+      "step": 485
+    },
+    {
+      "epoch": 0.02478281966721989,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0019496000000000001,
+      "loss": 3.0209,
+      "step": 486
+    },
+    {
+      "epoch": 0.024833813123325284,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0019532,
+      "loss": 3.027,
+      "step": 487
+    },
+    {
+      "epoch": 0.024884806579430672,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0019568,
+      "loss": 3.0356,
+      "step": 488
+    },
+    {
+      "epoch": 0.024935800035536065,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0019604,
+      "loss": 3.0062,
+      "step": 489
+    },
+    {
+      "epoch": 0.024986793491641457,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001964,
+      "loss": 3.018,
+      "step": 490
+    },
+    {
+      "epoch": 0.025037786947746846,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0019676,
+      "loss": 3.0411,
+      "step": 491
+    },
+    {
+      "epoch": 0.025088780403852238,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0019712,
+      "loss": 2.9976,
+      "step": 492
+    },
+    {
+      "epoch": 0.025139773859957627,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0019748,
+      "loss": 2.998,
+      "step": 493
+    },
+    {
+      "epoch": 0.02519076731606302,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0019784,
+      "loss": 3.0402,
+      "step": 494
+    },
+    {
+      "epoch": 0.025241760772168408,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.001982,
+      "loss": 3.0316,
+      "step": 495
+    },
+    {
+      "epoch": 0.0252927542282738,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0019856,
+      "loss": 3.0415,
+      "step": 496
+    },
+    {
+      "epoch": 0.025343747684379192,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0019892,
+      "loss": 3.007,
+      "step": 497
+    },
+    {
+      "epoch": 0.02539474114048458,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0019928,
+      "loss": 3.0443,
+      "step": 498
+    },
+    {
+      "epoch": 0.025445734596589973,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0019964,
+      "loss": 3.0115,
+      "step": 499
+    },
+    {
+      "epoch": 0.02549672805269536,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 3.0276,
+      "step": 500
+    },
+    {
+      "epoch": 0.025547721508800754,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 3.0394,
+      "step": 501
+    },
+    {
+      "epoch": 0.025598714964906143,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 3.001,
+      "step": 502
+    },
+    {
+      "epoch": 0.025649708421011535,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 3.0174,
+      "step": 503
+    },
+    {
+      "epoch": 0.025700701877116927,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 3.0325,
+      "step": 504
+    },
+    {
+      "epoch": 0.025751695333222316,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 3.0238,
+      "step": 505
+    },
+    {
+      "epoch": 0.025802688789327708,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 3.0291,
+      "step": 506
+    },
+    {
+      "epoch": 0.025853682245433097,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 3.003,
+      "step": 507
+    },
+    {
+      "epoch": 0.02590467570153849,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 3.0179,
+      "step": 508
+    },
+    {
+      "epoch": 0.025955669157643878,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 3.0096,
+      "step": 509
+    },
+    {
+      "epoch": 0.02600666261374927,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 3.0276,
+      "step": 510
+    },
+    {
+      "epoch": 0.026057656069854662,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 3.0291,
+      "step": 511
+    },
+    {
+      "epoch": 0.02610864952596005,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 3.041,
+      "step": 512
+    },
+    {
+      "epoch": 0.026159642982065443,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 3.0498,
+      "step": 513
+    },
+    {
+      "epoch": 0.026210636438170832,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 2.9924,
+      "step": 514
+    },
+    {
+      "epoch": 0.026261629894276224,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 2.9952,
+      "step": 515
+    },
+    {
+      "epoch": 0.026312623350381613,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 3.0308,
+      "step": 516
+    },
+    {
+      "epoch": 0.026363616806487005,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 3.0024,
+      "step": 517
+    },
+    {
+      "epoch": 0.026414610262592397,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 3.0411,
+      "step": 518
+    },
+    {
+      "epoch": 0.026465603718697786,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.0144,
+      "step": 519
+    },
+    {
+      "epoch": 0.02651659717480318,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 3.0286,
+      "step": 520
+    },
+    {
+      "epoch": 0.026567590630908567,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.0217,
+      "step": 521
+    },
+    {
+      "epoch": 0.02661858408701396,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 3.0053,
+      "step": 522
+    },
+    {
+      "epoch": 0.026669577543119348,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 3.0125,
+      "step": 523
+    },
+    {
+      "epoch": 0.02672057099922474,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 3.0594,
+      "step": 524
+    },
+    {
+      "epoch": 0.026771564455330132,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.9919,
+      "step": 525
+    },
+    {
+      "epoch": 0.02682255791143552,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 3.0022,
+      "step": 526
+    },
+    {
+      "epoch": 0.026873551367540913,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 3.0472,
+      "step": 527
+    },
+    {
+      "epoch": 0.026924544823646302,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 2.9982,
+      "step": 528
+    },
+    {
+      "epoch": 0.026975538279751694,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 3.0257,
+      "step": 529
+    },
+    {
+      "epoch": 0.027026531735857083,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 3.0223,
+      "step": 530
+    },
+    {
+      "epoch": 0.027077525191962475,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 3.032,
+      "step": 531
+    },
+    {
+      "epoch": 0.027128518648067868,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 3.0151,
+      "step": 532
+    },
+    {
+      "epoch": 0.027179512104173256,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 3.0225,
+      "step": 533
+    },
+    {
+      "epoch": 0.02723050556027865,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 3.0194,
+      "step": 534
+    },
+    {
+      "epoch": 0.027281499016384037,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 3.0192,
+      "step": 535
+    },
+    {
+      "epoch": 0.02733249247248943,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 3.0088,
+      "step": 536
+    },
+    {
+      "epoch": 0.027383485928594818,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.9903,
+      "step": 537
+    },
+    {
+      "epoch": 0.02743447938470021,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 3.0409,
+      "step": 538
+    },
+    {
+      "epoch": 0.027485472840805603,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.9926,
+      "step": 539
+    },
+    {
+      "epoch": 0.02753646629691099,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 3.0315,
+      "step": 540
+    },
+    {
+      "epoch": 0.027587459753016384,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 3.0218,
+      "step": 541
+    },
+    {
+      "epoch": 0.027638453209121772,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 3.0216,
+      "step": 542
+    },
+    {
+      "epoch": 0.027689446665227165,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 3.0163,
+      "step": 543
+    },
+    {
+      "epoch": 0.027740440121332553,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 3.0016,
+      "step": 544
+    },
+    {
+      "epoch": 0.027791433577437945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 3.0291,
+      "step": 545
+    },
+    {
+      "epoch": 0.027842427033543338,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 3.0072,
+      "step": 546
+    },
+    {
+      "epoch": 0.027893420489648726,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 3.0227,
+      "step": 547
+    },
+    {
+      "epoch": 0.02794441394575412,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 3.054,
+      "step": 548
+    },
+    {
+      "epoch": 0.027995407401859507,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 3.0168,
+      "step": 549
+    },
+    {
+      "epoch": 0.0280464008579649,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 3.0068,
+      "step": 550
+    },
+    {
+      "epoch": 0.02809739431407029,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 3.0224,
+      "step": 551
+    },
+    {
+      "epoch": 0.02814838777017568,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 3.0354,
+      "step": 552
+    },
+    {
+      "epoch": 0.028199381226281073,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 3.02,
+      "step": 553
+    },
+    {
+      "epoch": 0.02825037468238646,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 3.029,
+      "step": 554
+    },
+    {
+      "epoch": 0.028301368138491854,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 3.0373,
+      "step": 555
+    },
+    {
+      "epoch": 0.028352361594597243,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 3.0104,
+      "step": 556
+    },
+    {
+      "epoch": 0.028403355050702635,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 3.0002,
+      "step": 557
+    },
+    {
+      "epoch": 0.028454348506808023,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 3.0184,
+      "step": 558
+    },
+    {
+      "epoch": 0.028505341962913416,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 2.9741,
+      "step": 559
+    },
+    {
+      "epoch": 0.028556335419018808,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 3.0235,
+      "step": 560
+    },
+    {
+      "epoch": 0.028607328875124197,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 3.0147,
+      "step": 561
+    },
+    {
+      "epoch": 0.02865832233122959,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.9888,
+      "step": 562
+    },
+    {
+      "epoch": 0.028709315787334978,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 3.0058,
+      "step": 563
+    },
+    {
+      "epoch": 0.02876030924344037,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9852,
+      "step": 564
+    },
+    {
+      "epoch": 0.02881130269954576,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 3.0039,
+      "step": 565
+    },
+    {
+      "epoch": 0.02886229615565115,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 3.0239,
+      "step": 566
+    },
+    {
+      "epoch": 0.028913289611756543,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 3.0111,
+      "step": 567
+    },
+    {
+      "epoch": 0.028964283067861932,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 3.0096,
+      "step": 568
+    },
+    {
+      "epoch": 0.029015276523967324,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 3.0078,
+      "step": 569
+    },
+    {
+      "epoch": 0.029066269980072713,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 3.034,
+      "step": 570
+    },
+    {
+      "epoch": 0.029117263436178105,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.9947,
+      "step": 571
+    },
+    {
+      "epoch": 0.029168256892283494,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9967,
+      "step": 572
+    },
+    {
+      "epoch": 0.029219250348388886,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 3.0163,
+      "step": 573
+    },
+    {
+      "epoch": 0.029270243804494278,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 3.0067,
+      "step": 574
+    },
+    {
+      "epoch": 0.029321237260599667,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 3.0046,
+      "step": 575
+    },
+    {
+      "epoch": 0.02937223071670506,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 3.0115,
+      "step": 576
+    },
+    {
+      "epoch": 0.029423224172810448,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 3.017,
+      "step": 577
+    },
+    {
+      "epoch": 0.02947421762891584,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 2.9961,
+      "step": 578
+    },
+    {
+      "epoch": 0.02952521108502123,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 3.0117,
+      "step": 579
+    },
+    {
+      "epoch": 0.02957620454112662,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9806,
+      "step": 580
+    },
+    {
+      "epoch": 0.029627197997232013,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 2.9997,
+      "step": 581
+    },
+    {
+      "epoch": 0.029678191453337402,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 3.0187,
+      "step": 582
+    },
+    {
+      "epoch": 0.029729184909442794,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 3.0028,
+      "step": 583
+    },
+    {
+      "epoch": 0.029780178365548183,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 2.9959,
+      "step": 584
+    },
+    {
+      "epoch": 0.029831171821653575,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 3.0243,
+      "step": 585
+    },
+    {
+      "epoch": 0.029882165277758964,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 3.009,
+      "step": 586
+    },
+    {
+      "epoch": 0.029933158733864356,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 3.0084,
+      "step": 587
+    },
+    {
+      "epoch": 0.029984152189969745,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 3.0264,
+      "step": 588
+    },
+    {
+      "epoch": 0.030035145646075137,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 3.028,
+      "step": 589
+    },
+    {
+      "epoch": 0.03008613910218053,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 3.0019,
+      "step": 590
+    },
+    {
+      "epoch": 0.030137132558285918,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 3.0206,
+      "step": 591
+    },
+    {
+      "epoch": 0.03018812601439131,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 3.0156,
+      "step": 592
+    },
+    {
+      "epoch": 0.0302391194704967,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 3.0324,
+      "step": 593
+    },
+    {
+      "epoch": 0.03029011292660209,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 3.001,
+      "step": 594
+    },
+    {
+      "epoch": 0.03034110638270748,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.9782,
+      "step": 595
+    },
+    {
+      "epoch": 0.030392099838812872,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 3.0081,
+      "step": 596
+    },
+    {
+      "epoch": 0.030443093294918264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 3.0211,
+      "step": 597
+    },
+    {
+      "epoch": 0.030494086751023653,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 3.0025,
+      "step": 598
+    },
+    {
+      "epoch": 0.030545080207129045,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 2.9747,
+      "step": 599
+    },
+    {
+      "epoch": 0.030596073663234434,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 2.9975,
+      "step": 600
+    },
+    {
+      "epoch": 0.030647067119339826,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.9839,
+      "step": 601
+    },
+    {
+      "epoch": 0.030698060575445215,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 3.0133,
+      "step": 602
+    },
+    {
+      "epoch": 0.030749054031550607,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 3.0249,
+      "step": 603
+    },
+    {
+      "epoch": 0.030800047487656,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 3.0066,
+      "step": 604
+    },
+    {
+      "epoch": 0.030851040943761388,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.9945,
+      "step": 605
+    },
+    {
+      "epoch": 0.03090203439986678,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 3.0142,
+      "step": 606
+    },
+    {
+      "epoch": 0.03095302785597217,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.0227,
+      "step": 607
+    },
+    {
+      "epoch": 0.03100402131207756,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9812,
+      "step": 608
+    },
+    {
+      "epoch": 0.03105501476818295,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9901,
+      "step": 609
+    },
+    {
+      "epoch": 0.031106008224288342,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9806,
+      "step": 610
+    },
+    {
+      "epoch": 0.031157001680393735,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 3.0247,
+      "step": 611
+    },
+    {
+      "epoch": 0.031207995136499123,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 3.0152,
+      "step": 612
+    },
+    {
+      "epoch": 0.031258988592604516,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9976,
+      "step": 613
+    },
+    {
+      "epoch": 0.031309982048709904,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 3.0231,
+      "step": 614
+    },
+    {
+      "epoch": 0.03136097550481529,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 3.0191,
+      "step": 615
+    },
+    {
+      "epoch": 0.03141196896092069,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 3.0179,
+      "step": 616
+    },
+    {
+      "epoch": 0.03146296241702608,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.9943,
+      "step": 617
+    },
+    {
+      "epoch": 0.031513955873131466,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9712,
+      "step": 618
+    },
+    {
+      "epoch": 0.03156494932923686,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 3.0246,
+      "step": 619
+    },
+    {
+      "epoch": 0.03161594278534225,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 3.0083,
+      "step": 620
+    },
+    {
+      "epoch": 0.03166693624144764,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9995,
+      "step": 621
+    },
+    {
+      "epoch": 0.03171792969755303,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 3.0046,
+      "step": 622
+    },
+    {
+      "epoch": 0.031768923153658424,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.9952,
+      "step": 623
+    },
+    {
+      "epoch": 0.03181991660976381,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.9884,
+      "step": 624
+    },
+    {
+      "epoch": 0.0318709100658692,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 3.0029,
+      "step": 625
+    },
+    {
+      "epoch": 0.0319219035219746,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.9836,
+      "step": 626
+    },
+    {
+      "epoch": 0.031972896978079986,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 3.0048,
+      "step": 627
+    },
+    {
+      "epoch": 0.032023890434185374,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9791,
+      "step": 628
+    },
+    {
+      "epoch": 0.03207488389029076,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.991,
+      "step": 629
+    },
+    {
+      "epoch": 0.03212587734639616,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.997,
+      "step": 630
+    },
+    {
+      "epoch": 0.03217687080250155,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 3.0027,
+      "step": 631
+    },
+    {
+      "epoch": 0.032227864258606936,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9589,
+      "step": 632
+    },
+    {
+      "epoch": 0.03227885771471233,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 3.0002,
+      "step": 633
+    },
+    {
+      "epoch": 0.03232985117081772,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9962,
+      "step": 634
+    },
+    {
+      "epoch": 0.03238084462692311,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 2.9916,
+      "step": 635
+    },
+    {
+      "epoch": 0.0324318380830285,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9702,
+      "step": 636
+    },
+    {
+      "epoch": 0.032482831539133894,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.9876,
+      "step": 637
+    },
+    {
+      "epoch": 0.03253382499523928,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 3.0072,
+      "step": 638
+    },
+    {
+      "epoch": 0.03258481845134467,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9875,
+      "step": 639
+    },
+    {
+      "epoch": 0.03263581190745007,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 3.0104,
+      "step": 640
+    },
+    {
+      "epoch": 0.032686805363555456,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 3.0244,
+      "step": 641
+    },
+    {
+      "epoch": 0.032737798819660845,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9955,
+      "step": 642
+    },
+    {
+      "epoch": 0.03278879227576623,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9763,
+      "step": 643
+    },
+    {
+      "epoch": 0.03283978573187163,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.9716,
+      "step": 644
+    },
+    {
+      "epoch": 0.03289077918797702,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 3.021,
+      "step": 645
+    },
+    {
+      "epoch": 0.03294177264408241,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9765,
+      "step": 646
+    },
+    {
+      "epoch": 0.0329927661001878,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 3.0125,
+      "step": 647
+    },
+    {
+      "epoch": 0.03304375955629319,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 3.0276,
+      "step": 648
+    },
+    {
+      "epoch": 0.03309475301239858,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 3.006,
+      "step": 649
+    },
+    {
+      "epoch": 0.03314574646850397,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 3.0018,
+      "step": 650
+    },
+    {
+      "epoch": 0.033196739924609364,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9886,
+      "step": 651
+    },
+    {
+      "epoch": 0.03324773338071475,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 3.0015,
+      "step": 652
+    },
+    {
+      "epoch": 0.03329872683682014,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.9968,
+      "step": 653
+    },
+    {
+      "epoch": 0.03334972029292554,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 2.9991,
+      "step": 654
+    },
+    {
+      "epoch": 0.033400713749030926,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.9923,
+      "step": 655
+    },
+    {
+      "epoch": 0.033451707205136315,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 3.0224,
+      "step": 656
+    },
+    {
+      "epoch": 0.033502700661241704,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9877,
+      "step": 657
+    },
+    {
+      "epoch": 0.0335536941173471,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 3.0096,
+      "step": 658
+    },
+    {
+      "epoch": 0.03360468757345249,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.993,
+      "step": 659
+    },
+    {
+      "epoch": 0.03365568102955788,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 3.0091,
+      "step": 660
+    },
+    {
+      "epoch": 0.03370667448566327,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 3.0029,
+      "step": 661
+    },
+    {
+      "epoch": 0.03375766794176866,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.9928,
+      "step": 662
+    },
+    {
+      "epoch": 0.03380866139787405,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9808,
+      "step": 663
+    },
+    {
+      "epoch": 0.03385965485397944,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9843,
+      "step": 664
+    },
+    {
+      "epoch": 0.033910648310084834,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9637,
+      "step": 665
+    },
+    {
+      "epoch": 0.03396164176619022,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9699,
+      "step": 666
+    },
+    {
+      "epoch": 0.03401263522229561,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9635,
+      "step": 667
+    },
+    {
+      "epoch": 0.03406362867840101,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.996,
+      "step": 668
+    },
+    {
+      "epoch": 0.034114622134506396,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9992,
+      "step": 669
+    },
+    {
+      "epoch": 0.034165615590611785,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 3.0015,
+      "step": 670
+    },
+    {
+      "epoch": 0.034216609046717174,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.9968,
+      "step": 671
+    },
+    {
+      "epoch": 0.03426760250282257,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9897,
+      "step": 672
+    },
+    {
+      "epoch": 0.03431859595892796,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.996,
+      "step": 673
+    },
+    {
+      "epoch": 0.03436958941503335,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.0093,
+      "step": 674
+    },
+    {
+      "epoch": 0.03442058287113874,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9939,
+      "step": 675
+    },
+    {
+      "epoch": 0.03447157632724413,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9976,
+      "step": 676
+    },
+    {
+      "epoch": 0.03452256978334952,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9666,
+      "step": 677
+    },
+    {
+      "epoch": 0.03457356323945491,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9829,
+      "step": 678
+    },
+    {
+      "epoch": 0.034624556695560305,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9838,
+      "step": 679
+    },
+    {
+      "epoch": 0.03467555015166569,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9913,
+      "step": 680
+    },
+    {
+      "epoch": 0.03472654360777108,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 3.0017,
+      "step": 681
+    },
+    {
+      "epoch": 0.03477753706387648,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9956,
+      "step": 682
+    },
+    {
+      "epoch": 0.034828530519981866,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 3.0064,
+      "step": 683
+    },
+    {
+      "epoch": 0.034879523976087255,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9987,
+      "step": 684
+    },
+    {
+      "epoch": 0.034930517432192644,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9859,
+      "step": 685
+    },
+    {
+      "epoch": 0.03498151088829804,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.972,
+      "step": 686
+    },
+    {
+      "epoch": 0.03503250434440343,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9907,
+      "step": 687
+    },
+    {
+      "epoch": 0.03508349780050882,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9937,
+      "step": 688
+    },
+    {
+      "epoch": 0.03513449125661421,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9707,
+      "step": 689
+    },
+    {
+      "epoch": 0.0351854847127196,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.9778,
+      "step": 690
+    },
+    {
+      "epoch": 0.03523647816882499,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 2.9905,
+      "step": 691
+    },
+    {
+      "epoch": 0.03528747162493038,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 2.9855,
+      "step": 692
+    },
+    {
+      "epoch": 0.035338465081035775,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 2.9701,
+      "step": 693
+    },
+    {
+      "epoch": 0.035389458537141164,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 3.0313,
+      "step": 694
+    },
+    {
+      "epoch": 0.03544045199324655,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.977,
+      "step": 695
+    },
+    {
+      "epoch": 0.03549144544935195,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9876,
+      "step": 696
+    },
+    {
+      "epoch": 0.03554243890545734,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 3.0028,
+      "step": 697
+    },
+    {
+      "epoch": 0.035593432361562725,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9875,
+      "step": 698
+    },
+    {
+      "epoch": 0.035644425817668114,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9714,
+      "step": 699
+    },
+    {
+      "epoch": 0.03569541927377351,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9885,
+      "step": 700
+    },
+    {
+      "epoch": 0.0357464127298789,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.013,
+      "step": 701
+    },
+    {
+      "epoch": 0.03579740618598429,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9714,
+      "step": 702
+    },
+    {
+      "epoch": 0.03584839964208968,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9896,
+      "step": 703
+    },
+    {
+      "epoch": 0.03589939309819507,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9724,
+      "step": 704
+    },
+    {
+      "epoch": 0.03595038655430046,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9946,
+      "step": 705
+    },
+    {
+      "epoch": 0.03600138001040585,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.989,
+      "step": 706
+    },
+    {
+      "epoch": 0.036052373466511245,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9707,
+      "step": 707
+    },
+    {
+      "epoch": 0.036103366922616634,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9771,
+      "step": 708
+    },
+    {
+      "epoch": 0.03615436037872202,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 2.977,
+      "step": 709
+    },
+    {
+      "epoch": 0.03620535383482742,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9867,
+      "step": 710
+    },
+    {
+      "epoch": 0.03625634729093281,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9976,
+      "step": 711
+    },
+    {
+      "epoch": 0.036307340747038196,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.9954,
+      "step": 712
+    },
+    {
+      "epoch": 0.036358334203143584,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9855,
+      "step": 713
+    },
+    {
+      "epoch": 0.03640932765924898,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9898,
+      "step": 714
+    },
+    {
+      "epoch": 0.03646032111535437,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9762,
+      "step": 715
+    },
+    {
+      "epoch": 0.03651131457145976,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9678,
+      "step": 716
+    },
+    {
+      "epoch": 0.03656230802756515,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.99,
+      "step": 717
+    },
+    {
+      "epoch": 0.03661330148367054,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.9899,
+      "step": 718
+    },
+    {
+      "epoch": 0.03666429493977593,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.9954,
+      "step": 719
+    },
+    {
+      "epoch": 0.03671528839588132,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9695,
+      "step": 720
+    },
+    {
+      "epoch": 0.036766281851986715,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9871,
+      "step": 721
+    },
+    {
+      "epoch": 0.036817275308092104,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9964,
+      "step": 722
+    },
+    {
+      "epoch": 0.03686826876419749,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9841,
+      "step": 723
+    },
+    {
+      "epoch": 0.03691926222030289,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9894,
+      "step": 724
+    },
+    {
+      "epoch": 0.03697025567640828,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9842,
+      "step": 725
+    },
+    {
+      "epoch": 0.037021249132513666,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9811,
+      "step": 726
+    },
+    {
+      "epoch": 0.037072242588619055,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 3.0047,
+      "step": 727
+    },
+    {
+      "epoch": 0.03712323604472445,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 3.0201,
+      "step": 728
+    },
+    {
+      "epoch": 0.03717422950082984,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9794,
+      "step": 729
+    },
+    {
+      "epoch": 0.03722522295693523,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.979,
+      "step": 730
+    },
+    {
+      "epoch": 0.03727621641304062,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9864,
+      "step": 731
+    },
+    {
+      "epoch": 0.03732720986914601,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9995,
+      "step": 732
+    },
+    {
+      "epoch": 0.0373782033252514,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9767,
+      "step": 733
+    },
+    {
+      "epoch": 0.03742919678135679,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9764,
+      "step": 734
+    },
+    {
+      "epoch": 0.037480190237462185,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9846,
+      "step": 735
+    },
+    {
+      "epoch": 0.037531183693567574,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.976,
+      "step": 736
+    },
+    {
+      "epoch": 0.03758217714967296,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 3.0034,
+      "step": 737
+    },
+    {
+      "epoch": 0.03763317060577836,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.981,
+      "step": 738
+    },
+    {
+      "epoch": 0.03768416406188375,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9817,
+      "step": 739
+    },
+    {
+      "epoch": 0.037735157517989136,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9573,
+      "step": 740
+    },
+    {
+      "epoch": 0.037786150974094525,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.99,
+      "step": 741
+    },
+    {
+      "epoch": 0.03783714443019992,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 3.0065,
+      "step": 742
+    },
+    {
+      "epoch": 0.03788813788630531,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9913,
+      "step": 743
+    },
+    {
+      "epoch": 0.0379391313424107,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.994,
+      "step": 744
+    },
+    {
+      "epoch": 0.037990124798516094,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9708,
+      "step": 745
+    },
+    {
+      "epoch": 0.03804111825462148,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9791,
+      "step": 746
+    },
+    {
+      "epoch": 0.03809211171072687,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9952,
+      "step": 747
+    },
+    {
+      "epoch": 0.03814310516683226,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9773,
+      "step": 748
+    },
+    {
+      "epoch": 0.038194098622937656,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9794,
+      "step": 749
+    },
+    {
+      "epoch": 0.038245092079043044,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 2.9661,
+      "step": 750
+    },
+    {
+      "epoch": 0.03829608553514843,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.9814,
+      "step": 751
+    },
+    {
+      "epoch": 0.03834707899125383,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9738,
+      "step": 752
+    },
+    {
+      "epoch": 0.03839807244735922,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 3.0075,
+      "step": 753
+    },
+    {
+      "epoch": 0.038449065903464606,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9612,
+      "step": 754
+    },
+    {
+      "epoch": 0.038500059359569995,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9706,
+      "step": 755
+    },
+    {
+      "epoch": 0.03855105281567539,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9807,
+      "step": 756
+    },
+    {
+      "epoch": 0.03860204627178078,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9867,
+      "step": 757
+    },
+    {
+      "epoch": 0.03865303972788617,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9892,
+      "step": 758
+    },
+    {
+      "epoch": 0.038704033183991564,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9904,
+      "step": 759
+    },
+    {
+      "epoch": 0.03875502664009695,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9529,
+      "step": 760
+    },
+    {
+      "epoch": 0.03880602009620234,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9793,
+      "step": 761
+    },
+    {
+      "epoch": 0.03885701355230773,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9787,
+      "step": 762
+    },
+    {
+      "epoch": 0.038908007008413126,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.987,
+      "step": 763
+    },
+    {
+      "epoch": 0.038959000464518515,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9709,
+      "step": 764
+    },
+    {
+      "epoch": 0.0390099939206239,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.983,
+      "step": 765
+    },
+    {
+      "epoch": 0.0390609873767293,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9909,
+      "step": 766
+    },
+    {
+      "epoch": 0.03911198083283469,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9729,
+      "step": 767
+    },
+    {
+      "epoch": 0.039162974288940076,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9731,
+      "step": 768
+    },
+    {
+      "epoch": 0.039213967745045465,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 3.0146,
+      "step": 769
+    },
+    {
+      "epoch": 0.03926496120115086,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9607,
+      "step": 770
+    },
+    {
+      "epoch": 0.03931595465725625,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9716,
+      "step": 771
+    },
+    {
+      "epoch": 0.03936694811336164,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9886,
+      "step": 772
+    },
+    {
+      "epoch": 0.039417941569467034,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9885,
+      "step": 773
+    },
+    {
+      "epoch": 0.03946893502557242,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9708,
+      "step": 774
+    },
+    {
+      "epoch": 0.03951992848167781,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9531,
+      "step": 775
+    },
+    {
+      "epoch": 0.0395709219377832,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9786,
+      "step": 776
+    },
+    {
+      "epoch": 0.039621915393888596,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9579,
+      "step": 777
+    },
+    {
+      "epoch": 0.039672908849993985,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 3.0027,
+      "step": 778
+    },
+    {
+      "epoch": 0.03972390230609937,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9959,
+      "step": 779
+    },
+    {
+      "epoch": 0.03977489576220477,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.983,
+      "step": 780
+    },
+    {
+      "epoch": 0.03982588921831016,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9689,
+      "step": 781
+    },
+    {
+      "epoch": 0.03987688267441555,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 782
+    },
+    {
+      "epoch": 0.039927876130520935,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.991,
+      "step": 783
+    },
+    {
+      "epoch": 0.03997886958662633,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9795,
+      "step": 784
+    },
+    {
+      "epoch": 0.04002986304273172,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9668,
+      "step": 785
+    },
+    {
+      "epoch": 0.04008085649883711,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.9844,
+      "step": 786
+    },
+    {
+      "epoch": 0.040131849954942504,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9725,
+      "step": 787
+    },
+    {
+      "epoch": 0.04018284341104789,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.9671,
+      "step": 788
+    },
+    {
+      "epoch": 0.04023383686715328,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9628,
+      "step": 789
+    },
+    {
+      "epoch": 0.04028483032325867,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9689,
+      "step": 790
+    },
+    {
+      "epoch": 0.040335823779364066,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9611,
+      "step": 791
+    },
+    {
+      "epoch": 0.040386817235469455,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.98,
+      "step": 792
+    },
+    {
+      "epoch": 0.040437810691574844,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9698,
+      "step": 793
+    },
+    {
+      "epoch": 0.04048880414768024,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9954,
+      "step": 794
+    },
+    {
+      "epoch": 0.04053979760378563,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.978,
+      "step": 795
+    },
+    {
+      "epoch": 0.04059079105989102,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9738,
+      "step": 796
+    },
+    {
+      "epoch": 0.040641784515996406,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9509,
+      "step": 797
+    },
+    {
+      "epoch": 0.0406927779721018,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9655,
+      "step": 798
+    },
+    {
+      "epoch": 0.04074377142820719,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9791,
+      "step": 799
+    },
+    {
+      "epoch": 0.04079476488431258,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9535,
+      "step": 800
+    },
+    {
+      "epoch": 0.040845758340417974,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9883,
+      "step": 801
+    },
+    {
+      "epoch": 0.04089675179652336,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9853,
+      "step": 802
+    },
+    {
+      "epoch": 0.04094774525262875,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 803
+    },
+    {
+      "epoch": 0.04099873870873414,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9845,
+      "step": 804
+    },
+    {
+      "epoch": 0.041049732164839536,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9864,
+      "step": 805
+    },
+    {
+      "epoch": 0.041100725620944925,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.977,
+      "step": 806
+    },
+    {
+      "epoch": 0.041151719077050314,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 2.9736,
+      "step": 807
+    },
+    {
+      "epoch": 0.04120271253315571,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.9684,
+      "step": 808
+    },
+    {
+      "epoch": 0.0412537059892611,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.9934,
+      "step": 809
+    },
+    {
+      "epoch": 0.04130469944536649,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.9661,
+      "step": 810
+    },
+    {
+      "epoch": 0.041355692901471876,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.9525,
+      "step": 811
+    },
+    {
+      "epoch": 0.04140668635757727,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9723,
+      "step": 812
+    },
+    {
+      "epoch": 0.04145767981368266,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.981,
+      "step": 813
+    },
+    {
+      "epoch": 0.04150867326978805,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9664,
+      "step": 814
+    },
+    {
+      "epoch": 0.041559666725893445,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9441,
+      "step": 815
+    },
+    {
+      "epoch": 0.04161066018199883,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9583,
+      "step": 816
+    },
+    {
+      "epoch": 0.04166165363810422,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9814,
+      "step": 817
+    },
+    {
+      "epoch": 0.04171264709420961,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9696,
+      "step": 818
+    },
+    {
+      "epoch": 0.04176364055031501,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.9889,
+      "step": 819
+    },
+    {
+      "epoch": 0.041814634006420395,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9779,
+      "step": 820
+    },
+    {
+      "epoch": 0.041865627462525784,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9767,
+      "step": 821
+    },
+    {
+      "epoch": 0.04191662091863118,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9773,
+      "step": 822
+    },
+    {
+      "epoch": 0.04196761437473657,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9665,
+      "step": 823
+    },
+    {
+      "epoch": 0.04201860783084196,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9581,
+      "step": 824
+    },
+    {
+      "epoch": 0.042069601286947346,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9909,
+      "step": 825
+    },
+    {
+      "epoch": 0.04212059474305274,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9634,
+      "step": 826
+    },
+    {
+      "epoch": 0.04217158819915813,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9481,
+      "step": 827
+    },
+    {
+      "epoch": 0.04222258165526352,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9728,
+      "step": 828
+    },
+    {
+      "epoch": 0.042273575111368915,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9473,
+      "step": 829
+    },
+    {
+      "epoch": 0.042324568567474304,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9894,
+      "step": 830
+    },
+    {
+      "epoch": 0.04237556202357969,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9683,
+      "step": 831
+    },
+    {
+      "epoch": 0.04242655547968508,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9534,
+      "step": 832
+    },
+    {
+      "epoch": 0.04247754893579048,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9671,
+      "step": 833
+    },
+    {
+      "epoch": 0.042528542391895865,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9656,
+      "step": 834
+    },
+    {
+      "epoch": 0.042579535848001254,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9581,
+      "step": 835
+    },
+    {
+      "epoch": 0.04263052930410664,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9772,
+      "step": 836
+    },
+    {
+      "epoch": 0.04268152276021204,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9874,
+      "step": 837
+    },
+    {
+      "epoch": 0.04273251621631743,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9787,
+      "step": 838
+    },
+    {
+      "epoch": 0.042783509672422816,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9538,
+      "step": 839
+    },
+    {
+      "epoch": 0.04283450312852821,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.963,
+      "step": 840
+    },
+    {
+      "epoch": 0.0428854965846336,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9722,
+      "step": 841
+    },
+    {
+      "epoch": 0.04293649004073899,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9511,
+      "step": 842
+    },
+    {
+      "epoch": 0.04298748349684438,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9836,
+      "step": 843
+    },
+    {
+      "epoch": 0.043038476952949774,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9819,
+      "step": 844
+    },
+    {
+      "epoch": 0.04308947040905516,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9605,
+      "step": 845
+    },
+    {
+      "epoch": 0.04314046386516055,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9659,
+      "step": 846
+    },
+    {
+      "epoch": 0.04319145732126595,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9622,
+      "step": 847
+    },
+    {
+      "epoch": 0.043242450777371336,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9803,
+      "step": 848
+    },
+    {
+      "epoch": 0.043293444233476724,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9653,
+      "step": 849
+    },
+    {
+      "epoch": 0.04334443768958211,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9512,
+      "step": 850
+    },
+    {
+      "epoch": 0.04339543114568751,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9677,
+      "step": 851
+    },
+    {
+      "epoch": 0.0434464246017929,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9954,
+      "step": 852
+    },
+    {
+      "epoch": 0.043497418057898286,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9599,
+      "step": 853
+    },
+    {
+      "epoch": 0.04354841151400368,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9635,
+      "step": 854
+    },
+    {
+      "epoch": 0.04359940497010907,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9786,
+      "step": 855
+    },
+    {
+      "epoch": 0.04365039842621446,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9409,
+      "step": 856
+    },
+    {
+      "epoch": 0.04370139188231985,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9543,
+      "step": 857
+    },
+    {
+      "epoch": 0.043752385338425244,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9893,
+      "step": 858
+    },
+    {
+      "epoch": 0.04380337879453063,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9621,
+      "step": 859
+    },
+    {
+      "epoch": 0.04385437225063602,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9641,
+      "step": 860
+    },
+    {
+      "epoch": 0.04390536570674142,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 3.0018,
+      "step": 861
+    },
+    {
+      "epoch": 0.043956359162846806,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.967,
+      "step": 862
+    },
+    {
+      "epoch": 0.044007352618952195,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9702,
+      "step": 863
+    },
+    {
+      "epoch": 0.04405834607505758,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.962,
+      "step": 864
+    },
+    {
+      "epoch": 0.04410933953116298,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9589,
+      "step": 865
+    },
+    {
+      "epoch": 0.04416033298726837,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.966,
+      "step": 866
+    },
+    {
+      "epoch": 0.04421132644337376,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9573,
+      "step": 867
+    },
+    {
+      "epoch": 0.04426231989947915,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9552,
+      "step": 868
+    },
+    {
+      "epoch": 0.04431331335558454,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9842,
+      "step": 869
+    },
+    {
+      "epoch": 0.04436430681168993,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9994,
+      "step": 870
+    },
+    {
+      "epoch": 0.04441530026779532,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9868,
+      "step": 871
+    },
+    {
+      "epoch": 0.044466293723900714,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.974,
+      "step": 872
+    },
+    {
+      "epoch": 0.0445172871800061,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9518,
+      "step": 873
+    },
+    {
+      "epoch": 0.04456828063611149,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9974,
+      "step": 874
+    },
+    {
+      "epoch": 0.04461927409221689,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9946,
+      "step": 875
+    },
+    {
+      "epoch": 0.044670267548322276,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.976,
+      "step": 876
+    },
+    {
+      "epoch": 0.044721261004427665,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 3.0037,
+      "step": 877
+    },
+    {
+      "epoch": 0.044772254460533054,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9859,
+      "step": 878
+    },
+    {
+      "epoch": 0.04482324791663845,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9612,
+      "step": 879
+    },
+    {
+      "epoch": 0.04487424137274384,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9644,
+      "step": 880
+    },
+    {
+      "epoch": 0.04492523482884923,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9748,
+      "step": 881
+    },
+    {
+      "epoch": 0.04497622828495462,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9847,
+      "step": 882
+    },
+    {
+      "epoch": 0.04502722174106001,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9562,
+      "step": 883
+    },
+    {
+      "epoch": 0.0450782151971654,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9379,
+      "step": 884
+    },
+    {
+      "epoch": 0.04512920865327079,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9556,
+      "step": 885
+    },
+    {
+      "epoch": 0.045180202109376184,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9914,
+      "step": 886
+    },
+    {
+      "epoch": 0.04523119556548157,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9656,
+      "step": 887
+    },
+    {
+      "epoch": 0.04528218902158696,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9589,
+      "step": 888
+    },
+    {
+      "epoch": 0.04533318247769236,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9643,
+      "step": 889
+    },
+    {
+      "epoch": 0.045384175933797746,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9937,
+      "step": 890
+    },
+    {
+      "epoch": 0.045435169389903135,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.982,
+      "step": 891
+    },
+    {
+      "epoch": 0.045486162846008524,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9629,
+      "step": 892
+    },
+    {
+      "epoch": 0.04553715630211392,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9783,
+      "step": 893
+    },
+    {
+      "epoch": 0.04558814975821931,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9723,
+      "step": 894
+    },
+    {
+      "epoch": 0.0456391432143247,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9624,
+      "step": 895
+    },
+    {
+      "epoch": 0.04569013667043009,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.96,
+      "step": 896
+    },
+    {
+      "epoch": 0.04574113012653548,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9623,
+      "step": 897
+    },
+    {
+      "epoch": 0.04579212358264087,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9566,
+      "step": 898
+    },
+    {
+      "epoch": 0.04584311703874626,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9812,
+      "step": 899
+    },
+    {
+      "epoch": 0.045894110494851655,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9761,
+      "step": 900
+    },
+    {
+      "epoch": 0.04594510395095704,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9399,
+      "step": 901
+    },
+    {
+      "epoch": 0.04599609740706243,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9781,
+      "step": 902
+    },
+    {
+      "epoch": 0.04604709086316783,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9557,
+      "step": 903
+    },
+    {
+      "epoch": 0.046098084319273216,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9371,
+      "step": 904
+    },
+    {
+      "epoch": 0.046149077775378605,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.969,
+      "step": 905
+    },
+    {
+      "epoch": 0.046200071231483994,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9451,
+      "step": 906
+    },
+    {
+      "epoch": 0.04625106468758939,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9786,
+      "step": 907
+    },
+    {
+      "epoch": 0.04630205814369478,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9849,
+      "step": 908
+    },
+    {
+      "epoch": 0.04635305159980017,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9836,
+      "step": 909
+    },
+    {
+      "epoch": 0.04640404505590556,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9396,
+      "step": 910
+    },
+    {
+      "epoch": 0.04645503851201095,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9602,
+      "step": 911
+    },
+    {
+      "epoch": 0.04650603196811634,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9892,
+      "step": 912
+    },
+    {
+      "epoch": 0.04655702542422173,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9829,
+      "step": 913
+    },
+    {
+      "epoch": 0.046608018880327125,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9821,
+      "step": 914
+    },
+    {
+      "epoch": 0.046659012336432514,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9351,
+      "step": 915
+    },
+    {
+      "epoch": 0.0467100057925379,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9723,
+      "step": 916
+    },
+    {
+      "epoch": 0.0467609992486433,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9586,
+      "step": 917
+    },
+    {
+      "epoch": 0.04681199270474869,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9853,
+      "step": 918
+    },
+    {
+      "epoch": 0.046862986160854075,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9686,
+      "step": 919
+    },
+    {
+      "epoch": 0.046913979616959464,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9355,
+      "step": 920
+    },
+    {
+      "epoch": 0.04696497307306486,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9482,
+      "step": 921
+    },
+    {
+      "epoch": 0.04701596652917025,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9512,
+      "step": 922
+    },
+    {
+      "epoch": 0.04706695998527564,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9635,
+      "step": 923
+    },
+    {
+      "epoch": 0.04711795344138103,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9657,
+      "step": 924
+    },
+    {
+      "epoch": 0.04716894689748642,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9609,
+      "step": 925
+    },
+    {
+      "epoch": 0.04721994035359181,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9818,
+      "step": 926
+    },
+    {
+      "epoch": 0.0472709338096972,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9355,
+      "step": 927
+    },
+    {
+      "epoch": 0.047321927265802595,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9482,
+      "step": 928
+    },
+    {
+      "epoch": 0.047372920721907984,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9583,
+      "step": 929
+    },
+    {
+      "epoch": 0.04742391417801337,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9526,
+      "step": 930
+    },
+    {
+      "epoch": 0.04747490763411877,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9566,
+      "step": 931
+    },
+    {
+      "epoch": 0.04752590109022416,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9707,
+      "step": 932
+    },
+    {
+      "epoch": 0.047576894546329546,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9726,
+      "step": 933
+    },
+    {
+      "epoch": 0.047627888002434934,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9576,
+      "step": 934
+    },
+    {
+      "epoch": 0.04767888145854033,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9506,
+      "step": 935
+    },
+    {
+      "epoch": 0.04772987491464572,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9461,
+      "step": 936
+    },
+    {
+      "epoch": 0.04778086837075111,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9348,
+      "step": 937
+    },
+    {
+      "epoch": 0.0478318618268565,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9791,
+      "step": 938
+    },
+    {
+      "epoch": 0.04788285528296189,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9618,
+      "step": 939
+    },
+    {
+      "epoch": 0.04793384873906728,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 940
+    },
+    {
+      "epoch": 0.04798484219517267,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9395,
+      "step": 941
+    },
+    {
+      "epoch": 0.048035835651278065,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9578,
+      "step": 942
+    },
+    {
+      "epoch": 0.048086829107383454,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9657,
+      "step": 943
+    },
+    {
+      "epoch": 0.04813782256348884,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.943,
+      "step": 944
+    },
+    {
+      "epoch": 0.04818881601959424,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9736,
+      "step": 945
+    },
+    {
+      "epoch": 0.04823980947569963,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9387,
+      "step": 946
+    },
+    {
+      "epoch": 0.048290802931805016,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 947
+    },
+    {
+      "epoch": 0.048341796387910405,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9647,
+      "step": 948
+    },
+    {
+      "epoch": 0.0483927898440158,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9754,
+      "step": 949
+    },
+    {
+      "epoch": 0.04844378330012119,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9803,
+      "step": 950
+    },
+    {
+      "epoch": 0.04849477675622658,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9517,
+      "step": 951
+    },
+    {
+      "epoch": 0.04854577021233197,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9632,
+      "step": 952
+    },
+    {
+      "epoch": 0.04859676366843736,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.957,
+      "step": 953
+    },
+    {
+      "epoch": 0.04864775712454275,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.9598,
+      "step": 954
+    },
+    {
+      "epoch": 0.04869875058064814,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9632,
+      "step": 955
+    },
+    {
+      "epoch": 0.048749744036753535,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9523,
+      "step": 956
+    },
+    {
+      "epoch": 0.048800737492858924,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9638,
+      "step": 957
+    },
+    {
+      "epoch": 0.04885173094896431,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9492,
+      "step": 958
+    },
+    {
+      "epoch": 0.04890272440506971,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9592,
+      "step": 959
+    },
+    {
+      "epoch": 0.0489537178611751,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9808,
+      "step": 960
+    },
+    {
+      "epoch": 0.049004711317280486,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9682,
+      "step": 961
+    },
+    {
+      "epoch": 0.049055704773385875,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9501,
+      "step": 962
+    },
+    {
+      "epoch": 0.04910669822949127,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9528,
+      "step": 963
+    },
+    {
+      "epoch": 0.04915769168559666,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9684,
+      "step": 964
+    },
+    {
+      "epoch": 0.04920868514170205,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9463,
+      "step": 965
+    },
+    {
+      "epoch": 0.049259678597807444,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9531,
+      "step": 966
+    },
+    {
+      "epoch": 0.04931067205391283,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9606,
+      "step": 967
+    },
+    {
+      "epoch": 0.04936166551001822,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9696,
+      "step": 968
+    },
+    {
+      "epoch": 0.04941265896612361,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9584,
+      "step": 969
+    },
+    {
+      "epoch": 0.049463652422229006,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9525,
+      "step": 970
+    },
+    {
+      "epoch": 0.049514645878334394,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9513,
+      "step": 971
+    },
+    {
+      "epoch": 0.04956563933443978,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 3.008,
+      "step": 972
+    },
+    {
+      "epoch": 0.04961663279054518,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9575,
+      "step": 973
+    },
+    {
+      "epoch": 0.04966762624665057,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9367,
+      "step": 974
+    },
+    {
+      "epoch": 0.049718619702755956,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9858,
+      "step": 975
+    },
+    {
+      "epoch": 0.049769613158861345,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9396,
+      "step": 976
+    },
+    {
+      "epoch": 0.04982060661496674,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9321,
+      "step": 977
+    },
+    {
+      "epoch": 0.04987160007107213,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9269,
+      "step": 978
+    },
+    {
+      "epoch": 0.04992259352717752,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9527,
+      "step": 979
+    },
+    {
+      "epoch": 0.049973586983282914,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9505,
+      "step": 980
+    },
+    {
+      "epoch": 0.0500245804393883,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9815,
+      "step": 981
+    },
+    {
+      "epoch": 0.05007557389549369,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9697,
+      "step": 982
+    },
+    {
+      "epoch": 0.05012656735159908,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9407,
+      "step": 983
+    },
+    {
+      "epoch": 0.050177560807704476,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9709,
+      "step": 984
+    },
+    {
+      "epoch": 0.050228554263809864,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9359,
+      "step": 985
+    },
+    {
+      "epoch": 0.05027954771991525,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9553,
+      "step": 986
+    },
+    {
+      "epoch": 0.05033054117602065,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9555,
+      "step": 987
+    },
+    {
+      "epoch": 0.05038153463212604,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.934,
+      "step": 988
+    },
+    {
+      "epoch": 0.050432528088231426,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9573,
+      "step": 989
+    },
+    {
+      "epoch": 0.050483521544336815,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9387,
+      "step": 990
+    },
+    {
+      "epoch": 0.05053451500044221,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9657,
+      "step": 991
+    },
+    {
+      "epoch": 0.0505855084565476,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9517,
+      "step": 992
+    },
+    {
+      "epoch": 0.05063650191265299,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9918,
+      "step": 993
+    },
+    {
+      "epoch": 0.050687495368758384,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9327,
+      "step": 994
+    },
+    {
+      "epoch": 0.05073848882486377,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 995
+    },
+    {
+      "epoch": 0.05078948228096916,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9516,
+      "step": 996
+    },
+    {
+      "epoch": 0.05084047573707455,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9359,
+      "step": 997
+    },
+    {
+      "epoch": 0.050891469193179946,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9564,
+      "step": 998
+    },
+    {
+      "epoch": 0.050942462649285335,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9432,
+      "step": 999
+    },
+    {
+      "epoch": 0.05099345610539072,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9115,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05104444956149612,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9443,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05109544301760151,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9422,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0511464364737069,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9366,
+      "step": 1003
+    },
+    {
+      "epoch": 0.051197429929812285,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.942,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05124842338591768,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.954,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05129941684202307,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9443,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05135041029812846,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9616,
+      "step": 1007
+    },
+    {
+      "epoch": 0.051401403754233854,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.967,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05145239721033924,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9583,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05150339066644463,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9262,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05155438412255002,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9518,
+      "step": 1011
+    },
+    {
+      "epoch": 0.051605377578655416,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9345,
+      "step": 1012
+    },
+    {
+      "epoch": 0.051656371034760805,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.949,
+      "step": 1013
+    },
+    {
+      "epoch": 0.051707364490866194,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9814,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05175835794697159,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.961,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05180935140307698,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9683,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05186034485918237,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9518,
+      "step": 1017
+    },
+    {
+      "epoch": 0.051911338315287756,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9402,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05196233177139315,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9437,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05201332522749854,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9463,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05206431868360393,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9744,
+      "step": 1021
+    },
+    {
+      "epoch": 0.052115312139709324,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9624,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05216630559581471,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9479,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0522172990519201,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9485,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05226829250802549,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 2.9383,
+      "step": 1025
+    },
+    {
+      "epoch": 0.052319285964130886,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 2.9542,
+      "step": 1026
+    },
+    {
+      "epoch": 0.052370279420236275,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 2.9378,
+      "step": 1027
+    },
+    {
+      "epoch": 0.052421272876341664,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 2.9654,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05247226633244706,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.9771,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05252325978855245,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.9647,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05257425324465784,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9521,
+      "step": 1031
+    },
+    {
+      "epoch": 0.052625246700763226,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9291,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05267624015686862,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9461,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05272723361297401,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9633,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0527782270690794,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9482,
+      "step": 1035
+    },
+    {
+      "epoch": 0.052829220525184795,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9352,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05288021398129018,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9685,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05293120743739557,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.978,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05298220089350096,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9493,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05303319434960636,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9471,
+      "step": 1040
+    },
+    {
+      "epoch": 0.053084187805711745,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9551,
+      "step": 1041
+    },
+    {
+      "epoch": 0.053135181261817134,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.948,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05318617471792253,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9612,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05323716817402792,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9499,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05328816163013331,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9771,
+      "step": 1045
+    },
+    {
+      "epoch": 0.053339155086238696,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9177,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05339014854234409,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9279,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05344114199844948,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9555,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05349213545455487,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9421,
+      "step": 1049
+    },
+    {
+      "epoch": 0.053543128910660265,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9373,
+      "step": 1050
+    },
+    {
+      "epoch": 0.053594122366765654,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9532,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05364511582287104,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9463,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05369610927897643,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9647,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05374710273508183,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9558,
+      "step": 1054
+    },
+    {
+      "epoch": 0.053798096191187215,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9461,
+      "step": 1055
+    },
+    {
+      "epoch": 0.053849089647292604,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9566,
+      "step": 1056
+    },
+    {
+      "epoch": 0.053900083103398,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9393,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05395107655950339,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9573,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05400207001560878,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9496,
+      "step": 1059
+    },
+    {
+      "epoch": 0.054053063471714166,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9473,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05410405692781956,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9464,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05415505038392495,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9451,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05420604384003034,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9732,
+      "step": 1063
+    },
+    {
+      "epoch": 0.054257037296135735,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9486,
+      "step": 1064
+    },
+    {
+      "epoch": 0.054308030752241124,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9489,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05435902420834651,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9365,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0544100176644519,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9603,
+      "step": 1067
+    },
+    {
+      "epoch": 0.0544610111205573,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9663,
+      "step": 1068
+    },
+    {
+      "epoch": 0.054512004576662686,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9804,
+      "step": 1069
+    },
+    {
+      "epoch": 0.054562998032768074,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9414,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05461399148887347,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9485,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05466498494497886,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9399,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05471597840108425,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9658,
+      "step": 1073
+    },
+    {
+      "epoch": 0.054766971857189636,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9582,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05481796531329503,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9349,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05486895876940042,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9444,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05491995222550581,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9614,
+      "step": 1077
+    },
+    {
+      "epoch": 0.054970945681611205,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9361,
+      "step": 1078
+    },
+    {
+      "epoch": 0.055021939137716594,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9438,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05507293259382198,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9329,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05512392604992737,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9181,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05517491950603277,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.915,
+      "step": 1082
+    },
+    {
+      "epoch": 0.055225912962138156,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9196,
+      "step": 1083
+    },
+    {
+      "epoch": 0.055276906418243545,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9252,
+      "step": 1084
+    },
+    {
+      "epoch": 0.05532789987434894,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9644,
+      "step": 1085
+    },
+    {
+      "epoch": 0.05537889333045433,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9441,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05542988678655972,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9228,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05548088024266511,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0555318736987705,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.9474,
+      "step": 1089
+    },
+    {
+      "epoch": 0.05558286715487589,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9167,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05563386061098128,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9456,
+      "step": 1091
+    },
+    {
+      "epoch": 0.055684854067086675,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.914,
+      "step": 1092
+    },
+    {
+      "epoch": 0.055735847523192064,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.958,
+      "step": 1093
+    },
+    {
+      "epoch": 0.05578684097929745,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9643,
+      "step": 1094
+    },
+    {
+      "epoch": 0.05583783443540284,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9491,
+      "step": 1095
+    },
+    {
+      "epoch": 0.05588882789150824,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9741,
+      "step": 1096
+    },
+    {
+      "epoch": 0.055939821347613626,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9317,
+      "step": 1097
+    },
+    {
+      "epoch": 0.055990814803719015,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9476,
+      "step": 1098
+    },
+    {
+      "epoch": 0.05604180825982441,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9481,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0560928017159298,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9461,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05614379517203519,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9451,
+      "step": 1101
+    },
+    {
+      "epoch": 0.05619478862814058,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9586,
+      "step": 1102
+    },
+    {
+      "epoch": 0.05624578208424597,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9582,
+      "step": 1103
+    },
+    {
+      "epoch": 0.05629677554035136,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9424,
+      "step": 1104
+    },
+    {
+      "epoch": 0.05634776899645675,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9158,
+      "step": 1105
+    },
+    {
+      "epoch": 0.056398762452562146,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9557,
+      "step": 1106
+    },
+    {
+      "epoch": 0.056449755908667534,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9366,
+      "step": 1107
+    },
+    {
+      "epoch": 0.05650074936477292,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9654,
+      "step": 1108
+    },
+    {
+      "epoch": 0.05655174282087831,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9307,
+      "step": 1109
+    },
+    {
+      "epoch": 0.05660273627698371,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9681,
+      "step": 1110
+    },
+    {
+      "epoch": 0.056653729733089096,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9559,
+      "step": 1111
+    },
+    {
+      "epoch": 0.056704723189194485,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9448,
+      "step": 1112
+    },
+    {
+      "epoch": 0.05675571664529988,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9371,
+      "step": 1113
+    },
+    {
+      "epoch": 0.05680671010140527,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9526,
+      "step": 1114
+    },
+    {
+      "epoch": 0.05685770355751066,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9366,
+      "step": 1115
+    },
+    {
+      "epoch": 0.05690869701361605,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9316,
+      "step": 1116
+    },
+    {
+      "epoch": 0.05695969046972144,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9428,
+      "step": 1117
+    },
+    {
+      "epoch": 0.05701068392582683,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9585,
+      "step": 1118
+    },
+    {
+      "epoch": 0.05706167738193222,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9332,
+      "step": 1119
+    },
+    {
+      "epoch": 0.057112670838037616,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9556,
+      "step": 1120
+    },
+    {
+      "epoch": 0.057163664294143005,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9501,
+      "step": 1121
+    },
+    {
+      "epoch": 0.05721465775024839,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9616,
+      "step": 1122
+    },
+    {
+      "epoch": 0.05726565120635378,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9607,
+      "step": 1123
+    },
+    {
+      "epoch": 0.05731664466245918,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9603,
+      "step": 1124
+    },
+    {
+      "epoch": 0.057367638118564566,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9552,
+      "step": 1125
+    },
+    {
+      "epoch": 0.057418631574669955,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9292,
+      "step": 1126
+    },
+    {
+      "epoch": 0.05746962503077535,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9287,
+      "step": 1127
+    },
+    {
+      "epoch": 0.05752061848688074,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9696,
+      "step": 1128
+    },
+    {
+      "epoch": 0.05757161194298613,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9449,
+      "step": 1129
+    },
+    {
+      "epoch": 0.05762260539909152,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9375,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05767359885519691,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9495,
+      "step": 1131
+    },
+    {
+      "epoch": 0.0577245923113023,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9429,
+      "step": 1132
+    },
+    {
+      "epoch": 0.05777558576740769,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9137,
+      "step": 1133
+    },
+    {
+      "epoch": 0.057826579223513086,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9303,
+      "step": 1134
+    },
+    {
+      "epoch": 0.057877572679618475,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9573,
+      "step": 1135
+    },
+    {
+      "epoch": 0.057928566135723863,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9163,
+      "step": 1136
+    },
+    {
+      "epoch": 0.05797955959182925,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9599,
+      "step": 1137
+    },
+    {
+      "epoch": 0.05803055304793465,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9582,
+      "step": 1138
+    },
+    {
+      "epoch": 0.05808154650404004,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9396,
+      "step": 1139
+    },
+    {
+      "epoch": 0.058132539960145425,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9209,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05818353341625082,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9165,
+      "step": 1141
+    },
+    {
+      "epoch": 0.05823452687235621,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9459,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0582855203284616,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 1143
+    },
+    {
+      "epoch": 0.05833651378456699,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9674,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05838750724067238,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9291,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05843850069677777,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05848949415288316,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9421,
+      "step": 1147
+    },
+    {
+      "epoch": 0.058540487608988556,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9624,
+      "step": 1148
+    },
+    {
+      "epoch": 0.058591481065093945,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9238,
+      "step": 1149
+    },
+    {
+      "epoch": 0.058642474521199334,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9271,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05869346797730472,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9301,
+      "step": 1151
+    },
+    {
+      "epoch": 0.05874446143341012,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9656,
+      "step": 1152
+    },
+    {
+      "epoch": 0.05879545488951551,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9366,
+      "step": 1153
+    },
+    {
+      "epoch": 0.058846448345620896,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9658,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05889744180172629,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9457,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05894843525783168,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9505,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05899942871393707,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9493,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05905042217004246,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9424,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05910141562614785,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9364,
+      "step": 1159
+    },
+    {
+      "epoch": 0.05915240908225324,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9182,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05920340253835863,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9526,
+      "step": 1161
+    },
+    {
+      "epoch": 0.059254395994464026,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9279,
+      "step": 1162
+    },
+    {
+      "epoch": 0.059305389450569415,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.982,
+      "step": 1163
+    },
+    {
+      "epoch": 0.059356382906674804,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9416,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05940737636278019,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05945836981888559,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.949,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05950936327499098,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9319,
+      "step": 1167
+    },
+    {
+      "epoch": 0.059560356731096366,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9802,
+      "step": 1168
+    },
+    {
+      "epoch": 0.059611350187201755,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9362,
+      "step": 1169
+    },
+    {
+      "epoch": 0.05966234364330715,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9423,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05971333709941254,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9381,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05976433055551793,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9347,
+      "step": 1172
+    },
+    {
+      "epoch": 0.05981532401162332,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9349,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05986631746772871,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9323,
+      "step": 1174
+    },
+    {
+      "epoch": 0.0599173109238341,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9439,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05996830437993949,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9533,
+      "step": 1176
+    },
+    {
+      "epoch": 0.060019297836044885,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9558,
+      "step": 1177
+    },
+    {
+      "epoch": 0.060070291292150274,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9178,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06012128474825566,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9389,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06017227820436106,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9305,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06022327166046645,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9467,
+      "step": 1181
+    },
+    {
+      "epoch": 0.060274265116571836,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9414,
+      "step": 1182
+    },
+    {
+      "epoch": 0.060325258572677225,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9201,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06037625202878262,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.935,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06042724548488801,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.935,
+      "step": 1185
+    },
+    {
+      "epoch": 0.0604782389409934,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9671,
+      "step": 1186
+    },
+    {
+      "epoch": 0.060529232397098794,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9351,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06058022585320418,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9363,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06063121930930957,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9284,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06068221276541496,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9311,
+      "step": 1190
+    },
+    {
+      "epoch": 0.060733206221520356,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9177,
+      "step": 1191
+    },
+    {
+      "epoch": 0.060784199677625744,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9296,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06083519313373113,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9434,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06088618658983653,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9413,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06093718004594192,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9137,
+      "step": 1195
+    },
+    {
+      "epoch": 0.060988173502047306,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.923,
+      "step": 1196
+    },
+    {
+      "epoch": 0.061039166958152695,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9611,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06109016041425809,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.942,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06114115387036348,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9459,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06119214732646887,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9522,
+      "step": 1200
+    },
+    {
+      "epoch": 0.061243140782574264,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9341,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06129413423867965,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9432,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06134512769478504,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9372,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06139612115089043,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9191,
+      "step": 1204
+    },
+    {
+      "epoch": 0.061447114606995826,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9339,
+      "step": 1205
+    },
+    {
+      "epoch": 0.061498108063101214,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9294,
+      "step": 1206
+    },
+    {
+      "epoch": 0.0615491015192066,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 1207
+    },
+    {
+      "epoch": 0.061600094975312,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9495,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06165108843141739,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 3.015,
+      "step": 1209
+    },
+    {
+      "epoch": 0.061702081887522776,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9373,
+      "step": 1210
+    },
+    {
+      "epoch": 0.061753075343628165,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9248,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06180406879973356,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9118,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06185506225583895,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9605,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06190605571194434,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9451,
+      "step": 1214
+    },
+    {
+      "epoch": 0.061957049168049734,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9174,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06200804262415512,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9492,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06205903608026051,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9578,
+      "step": 1217
+    },
+    {
+      "epoch": 0.0621100295363659,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9359,
+      "step": 1218
+    },
+    {
+      "epoch": 0.062161022992471296,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9571,
+      "step": 1219
+    },
+    {
+      "epoch": 0.062212016448576685,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9572,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06226300990468207,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9448,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06231400336078747,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.962,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06236499681689286,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9272,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06241599027299825,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9319,
+      "step": 1224
+    },
+    {
+      "epoch": 0.062466983729103635,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.9474,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06251797718520903,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9246,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06256897064131442,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9216,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06261996409741981,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9244,
+      "step": 1228
+    },
+    {
+      "epoch": 0.0626709575535252,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9316,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06272195100963059,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9414,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06277294446573599,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9183,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06282393792184138,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9295,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06287493137794677,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9135,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06292592483405215,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06297691829015754,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9122,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06302791174626293,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06307890520236832,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9329,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06312989865847372,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9713,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06318089211457911,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9205,
+      "step": 1239
+    },
+    {
+      "epoch": 0.0632318855706845,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.926,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06328287902678989,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9308,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06333387248289528,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9211,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06338486593900067,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9233,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06343585939510606,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9451,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06348685285121146,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.935,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06353784630731685,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.952,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06358883976342224,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9482,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06363983321952763,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06369082667563301,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9242,
+      "step": 1249
+    },
+    {
+      "epoch": 0.0637418201317384,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9671,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06379281358784379,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9526,
+      "step": 1251
+    },
+    {
+      "epoch": 0.0638438070439492,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9204,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06389480050005458,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9223,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06394579395615997,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9271,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06399678741226536,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9402,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06404778086837075,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9208,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06409877432447614,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9364,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06414976778058153,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9247,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06420076123668693,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9537,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06425175469279232,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9365,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0643027481488977,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9209,
+      "step": 1261
+    },
+    {
+      "epoch": 0.0643537416050031,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9479,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06440473506110848,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9127,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06445572851721387,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9212,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06450672197331926,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9322,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06455771542942466,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.937,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06460870888553005,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9432,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06465970234163544,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9231,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06471069579774083,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9471,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06476168925384622,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9403,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06481268270995161,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9248,
+      "step": 1271
+    },
+    {
+      "epoch": 0.064863676166057,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.944,
+      "step": 1272
+    },
+    {
+      "epoch": 0.0649146696221624,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8961,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06496566307826779,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.941,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06501665653437318,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.9483,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06506764999047857,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.9493,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06511864344658395,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.9453,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06516963690268934,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9204,
+      "step": 1278
+    },
+    {
+      "epoch": 0.06522063035879473,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.9402,
+      "step": 1279
+    },
+    {
+      "epoch": 0.06527162381490013,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.9522,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06532261727100552,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1281
+    },
+    {
+      "epoch": 0.06537361072711091,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9346,
+      "step": 1282
+    },
+    {
+      "epoch": 0.0654246041832163,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9638,
+      "step": 1283
+    },
+    {
+      "epoch": 0.06547559763932169,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9265,
+      "step": 1284
+    },
+    {
+      "epoch": 0.06552659109542708,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.938,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06557758455153247,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.944,
+      "step": 1286
+    },
+    {
+      "epoch": 0.06562857800763787,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9472,
+      "step": 1287
+    },
+    {
+      "epoch": 0.06567957146374326,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9125,
+      "step": 1288
+    },
+    {
+      "epoch": 0.06573056491984865,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9194,
+      "step": 1289
+    },
+    {
+      "epoch": 0.06578155837595404,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9536,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06583255183205942,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9496,
+      "step": 1291
+    },
+    {
+      "epoch": 0.06588354528816481,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9211,
+      "step": 1292
+    },
+    {
+      "epoch": 0.0659345387442702,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9187,
+      "step": 1293
+    },
+    {
+      "epoch": 0.0659855322003756,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9507,
+      "step": 1294
+    },
+    {
+      "epoch": 0.066036525656481,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.943,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06608751911258638,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.911,
+      "step": 1296
+    },
+    {
+      "epoch": 0.06613851256869177,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9641,
+      "step": 1297
+    },
+    {
+      "epoch": 0.06618950602479716,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9496,
+      "step": 1298
+    },
+    {
+      "epoch": 0.06624049948090255,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9409,
+      "step": 1299
+    },
+    {
+      "epoch": 0.06629149293700794,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06634248639311334,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9242,
+      "step": 1301
+    },
+    {
+      "epoch": 0.06639347984921873,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9472,
+      "step": 1302
+    },
+    {
+      "epoch": 0.06644447330532412,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9278,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0664954667614295,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9075,
+      "step": 1304
+    },
+    {
+      "epoch": 0.0665464602175349,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9352,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06659745367364028,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.94,
+      "step": 1306
+    },
+    {
+      "epoch": 0.06664844712974567,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9592,
+      "step": 1307
+    },
+    {
+      "epoch": 0.06669944058585107,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9603,
+      "step": 1308
+    },
+    {
+      "epoch": 0.06675043404195646,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9175,
+      "step": 1309
+    },
+    {
+      "epoch": 0.06680142749806185,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9252,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06685242095416724,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9426,
+      "step": 1311
+    },
+    {
+      "epoch": 0.06690341441027263,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9525,
+      "step": 1312
+    },
+    {
+      "epoch": 0.06695440786637802,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9448,
+      "step": 1313
+    },
+    {
+      "epoch": 0.06700540132248341,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9545,
+      "step": 1314
+    },
+    {
+      "epoch": 0.06705639477858881,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8989,
+      "step": 1315
+    },
+    {
+      "epoch": 0.0671073882346942,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9329,
+      "step": 1316
+    },
+    {
+      "epoch": 0.06715838169079959,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9357,
+      "step": 1317
+    },
+    {
+      "epoch": 0.06720937514690498,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9295,
+      "step": 1318
+    },
+    {
+      "epoch": 0.06726036860301036,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9385,
+      "step": 1319
+    },
+    {
+      "epoch": 0.06731136205911575,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9197,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06736235551522114,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9207,
+      "step": 1321
+    },
+    {
+      "epoch": 0.06741334897132654,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9361,
+      "step": 1322
+    },
+    {
+      "epoch": 0.06746434242743193,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9277,
+      "step": 1323
+    },
+    {
+      "epoch": 0.06751533588353732,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9533,
+      "step": 1324
+    },
+    {
+      "epoch": 0.06756632933964271,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.942,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0676173227957481,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8926,
+      "step": 1326
+    },
+    {
+      "epoch": 0.06766831625185349,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9514,
+      "step": 1327
+    },
+    {
+      "epoch": 0.06771930970795888,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9183,
+      "step": 1328
+    },
+    {
+      "epoch": 0.06777030316406428,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9111,
+      "step": 1329
+    },
+    {
+      "epoch": 0.06782129662016967,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9262,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06787229007627506,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9355,
+      "step": 1331
+    },
+    {
+      "epoch": 0.06792328353238045,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8873,
+      "step": 1332
+    },
+    {
+      "epoch": 0.06797427698848583,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 1333
+    },
+    {
+      "epoch": 0.06802527044459122,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9426,
+      "step": 1334
+    },
+    {
+      "epoch": 0.06807626390069661,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9009,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06812725735680202,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9137,
+      "step": 1336
+    },
+    {
+      "epoch": 0.0681782508129074,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9411,
+      "step": 1337
+    },
+    {
+      "epoch": 0.06822924426901279,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9342,
+      "step": 1338
+    },
+    {
+      "epoch": 0.06828023772511818,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9316,
+      "step": 1339
+    },
+    {
+      "epoch": 0.06833123118122357,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9165,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06838222463732896,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9184,
+      "step": 1341
+    },
+    {
+      "epoch": 0.06843321809343435,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8909,
+      "step": 1342
+    },
+    {
+      "epoch": 0.06848421154953975,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9237,
+      "step": 1343
+    },
+    {
+      "epoch": 0.06853520500564514,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9431,
+      "step": 1344
+    },
+    {
+      "epoch": 0.06858619846175053,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9265,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06863719191785592,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.94,
+      "step": 1346
+    },
+    {
+      "epoch": 0.0686881853739613,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9037,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0687391788300667,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9133,
+      "step": 1348
+    },
+    {
+      "epoch": 0.06879017228617208,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9045,
+      "step": 1349
+    },
+    {
+      "epoch": 0.06884116574227749,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9381,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06889215919838287,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9316,
+      "step": 1351
+    },
+    {
+      "epoch": 0.06894315265448826,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9227,
+      "step": 1352
+    },
+    {
+      "epoch": 0.06899414611059365,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9209,
+      "step": 1353
+    },
+    {
+      "epoch": 0.06904513956669904,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9276,
+      "step": 1354
+    },
+    {
+      "epoch": 0.06909613302280443,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9404,
+      "step": 1355
+    },
+    {
+      "epoch": 0.06914712647890982,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9248,
+      "step": 1356
+    },
+    {
+      "epoch": 0.06919811993501522,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9405,
+      "step": 1357
+    },
+    {
+      "epoch": 0.06924911339112061,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8969,
+      "step": 1358
+    },
+    {
+      "epoch": 0.069300106847226,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9428,
+      "step": 1359
+    },
+    {
+      "epoch": 0.06935110030333139,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9201,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06940209375943678,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9366,
+      "step": 1361
+    },
+    {
+      "epoch": 0.06945308721554216,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9267,
+      "step": 1362
+    },
+    {
+      "epoch": 0.06950408067164755,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9546,
+      "step": 1363
+    },
+    {
+      "epoch": 0.06955507412775296,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9557,
+      "step": 1364
+    },
+    {
+      "epoch": 0.06960606758385834,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9297,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06965706103996373,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9438,
+      "step": 1366
+    },
+    {
+      "epoch": 0.06970805449606912,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9098,
+      "step": 1367
+    },
+    {
+      "epoch": 0.06975904795217451,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.953,
+      "step": 1368
+    },
+    {
+      "epoch": 0.0698100414082799,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9346,
+      "step": 1369
+    },
+    {
+      "epoch": 0.06986103486438529,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9185,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06991202832049069,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9272,
+      "step": 1371
+    },
+    {
+      "epoch": 0.06996302177659608,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.926,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07001401523270147,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9252,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07006500868880686,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9341,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07011600214491225,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9219,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07016699560101763,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8998,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07021798905712302,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.906,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07026898251322843,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9441,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07031997596933381,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9433,
+      "step": 1379
+    },
+    {
+      "epoch": 0.0703709694254392,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8956,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07042196288154459,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9427,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07047295633764998,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9422,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07052394979375537,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9582,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07057494324986076,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9202,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07062593670596616,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9425,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07067693016207155,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.917,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07072792361817694,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9449,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07077891707428233,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9277,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07082991053038772,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0708809039864931,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9567,
+      "step": 1390
+    },
+    {
+      "epoch": 0.0709318974425985,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9254,
+      "step": 1391
+    },
+    {
+      "epoch": 0.0709828908987039,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9107,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07103388435480928,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.91,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07108487781091467,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9326,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07113587126702006,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9244,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07118686472312545,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9205,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07123785817923084,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9158,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07128885163533623,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07133984509144163,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9214,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07139083854754702,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9138,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07144183200365241,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.924,
+      "step": 1401
+    },
+    {
+      "epoch": 0.0714928254597578,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9393,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07154381891586319,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9264,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07159481237196857,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9211,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07164580582807396,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9497,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07169679928417937,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9305,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07174779274028475,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9454,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07179878619639014,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.95,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07184977965249553,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9154,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07190077310860092,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9188,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07195176656470631,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9117,
+      "step": 1411
+    },
+    {
+      "epoch": 0.0720027600208117,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9508,
+      "step": 1412
+    },
+    {
+      "epoch": 0.0720537534769171,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9093,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07210474693302249,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.918,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07215574038912788,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9397,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07220673384523327,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9269,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07225772730133866,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9408,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07230872075744404,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9258,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07235971421354943,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9186,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07241070766965484,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9364,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07246170112576023,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9254,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07251269458186561,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.911,
+      "step": 1422
+    },
+    {
+      "epoch": 0.072563688037971,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.935,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07261468149407639,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9245,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07266567495018178,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.919,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07271666840628717,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9212,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07276766186239257,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9481,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07281865531849796,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9545,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07286964877460335,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9198,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07292064223070874,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9208,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07297163568681413,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9562,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07302262914291952,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8976,
+      "step": 1432
+    },
+    {
+      "epoch": 0.0730736225990249,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9364,
+      "step": 1433
+    },
+    {
+      "epoch": 0.0731246160551303,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9124,
+      "step": 1434
+    },
+    {
+      "epoch": 0.0731756095112357,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9029,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07322660296734108,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8998,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07327759642344647,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9099,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07332858987955186,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9297,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07337958333565725,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9415,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07343057679176264,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9273,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07348157024786804,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.9284,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07353256370397343,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9405,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07358355716007882,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9242,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07363455061618421,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9524,
+      "step": 1444
+    },
+    {
+      "epoch": 0.0736855440722896,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9092,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07373653752839499,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9289,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07378753098450037,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9195,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07383852444060578,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9264,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07388951789671117,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8919,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07394051135281655,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9492,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07399150480892194,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9373,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07404249826502733,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9336,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07409349172113272,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8914,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07414448517723811,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9172,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07419547863334351,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1455
+    },
+    {
+      "epoch": 0.0742464720894489,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.936,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07429746554555429,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.9059,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07434845900165968,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.9028,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07439945245776507,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9205,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07445044591387046,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.9406,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07450143936997584,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9196,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07455243282608125,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07460342628218664,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.932,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07465441973829202,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9172,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07470541319439741,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.928,
+      "step": 1465
+    },
+    {
+      "epoch": 0.0747564066505028,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9283,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07480740010660819,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8987,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07485839356271358,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.905,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07490938701881898,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9182,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07496038047492437,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9212,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07501137393102976,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9186,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07506236738713515,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9151,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07511336084324054,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9452,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07516435429934593,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.948,
+      "step": 1474
+    },
+    {
+      "epoch": 0.07521534775545131,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9381,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07526634121155672,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9243,
+      "step": 1476
+    },
+    {
+      "epoch": 0.0753173346676621,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9127,
+      "step": 1477
+    },
+    {
+      "epoch": 0.0753683281237675,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8975,
+      "step": 1478
+    },
+    {
+      "epoch": 0.07541932157987288,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9396,
+      "step": 1479
+    },
+    {
+      "epoch": 0.07547031503597827,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9127,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07552130849208366,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9235,
+      "step": 1481
+    },
+    {
+      "epoch": 0.07557230194818905,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.924,
+      "step": 1482
+    },
+    {
+      "epoch": 0.07562329540429445,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9387,
+      "step": 1483
+    },
+    {
+      "epoch": 0.07567428886039984,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9028,
+      "step": 1484
+    },
+    {
+      "epoch": 0.07572528231650523,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07577627577261062,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 1486
+    },
+    {
+      "epoch": 0.07582726922871601,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9017,
+      "step": 1487
+    },
+    {
+      "epoch": 0.0758782626848214,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9226,
+      "step": 1488
+    },
+    {
+      "epoch": 0.07592925614092678,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9096,
+      "step": 1489
+    },
+    {
+      "epoch": 0.07598024959703219,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9188,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07603124305313758,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9112,
+      "step": 1491
+    },
+    {
+      "epoch": 0.07608223650924296,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.928,
+      "step": 1492
+    },
+    {
+      "epoch": 0.07613322996534835,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9273,
+      "step": 1493
+    },
+    {
+      "epoch": 0.07618422342145374,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9365,
+      "step": 1494
+    },
+    {
+      "epoch": 0.07623521687755913,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07628621033366452,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9129,
+      "step": 1496
+    },
+    {
+      "epoch": 0.07633720378976992,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9154,
+      "step": 1497
+    },
+    {
+      "epoch": 0.07638819724587531,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9056,
+      "step": 1498
+    },
+    {
+      "epoch": 0.0764391907019807,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9098,
+      "step": 1499
+    },
+    {
+      "epoch": 0.07649018415808609,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9288,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07654117761419148,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9166,
+      "step": 1501
+    },
+    {
+      "epoch": 0.07659217107029687,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1502
+    },
+    {
+      "epoch": 0.07664316452640225,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.905,
+      "step": 1503
+    },
+    {
+      "epoch": 0.07669415798250766,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 1504
+    },
+    {
+      "epoch": 0.07674515143861305,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.946,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07679614489471843,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9113,
+      "step": 1506
+    },
+    {
+      "epoch": 0.07684713835082382,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 1507
+    },
+    {
+      "epoch": 0.07689813180692921,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9123,
+      "step": 1508
+    },
+    {
+      "epoch": 0.0769491252630346,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.922,
+      "step": 1509
+    },
+    {
+      "epoch": 0.07700011871913999,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9214,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07705111217524539,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9355,
+      "step": 1511
+    },
+    {
+      "epoch": 0.07710210563135078,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9122,
+      "step": 1512
+    },
+    {
+      "epoch": 0.07715309908745617,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9087,
+      "step": 1513
+    },
+    {
+      "epoch": 0.07720409254356156,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9284,
+      "step": 1514
+    },
+    {
+      "epoch": 0.07725508599966695,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9143,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07730607945577234,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9125,
+      "step": 1516
+    },
+    {
+      "epoch": 0.07735707291187773,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9208,
+      "step": 1517
+    },
+    {
+      "epoch": 0.07740806636798313,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8977,
+      "step": 1518
+    },
+    {
+      "epoch": 0.07745905982408852,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8868,
+      "step": 1519
+    },
+    {
+      "epoch": 0.0775100532801939,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9203,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0775610467362993,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9247,
+      "step": 1521
+    },
+    {
+      "epoch": 0.07761204019240468,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9136,
+      "step": 1522
+    },
+    {
+      "epoch": 0.07766303364851007,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9239,
+      "step": 1523
+    },
+    {
+      "epoch": 0.07771402710461546,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9196,
+      "step": 1524
+    },
+    {
+      "epoch": 0.07776502056072086,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.913,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07781601401682625,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8931,
+      "step": 1526
+    },
+    {
+      "epoch": 0.07786700747293164,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9382,
+      "step": 1527
+    },
+    {
+      "epoch": 0.07791800092903703,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.935,
+      "step": 1528
+    },
+    {
+      "epoch": 0.07796899438514242,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9321,
+      "step": 1529
+    },
+    {
+      "epoch": 0.0780199878412478,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0780709812973532,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.907,
+      "step": 1531
+    },
+    {
+      "epoch": 0.0781219747534586,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9005,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07817296820956399,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9031,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07822396166566938,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9266,
+      "step": 1534
+    },
+    {
+      "epoch": 0.07827495512177476,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.934,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07832594857788015,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9205,
+      "step": 1536
+    },
+    {
+      "epoch": 0.07837694203398554,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9411,
+      "step": 1537
+    },
+    {
+      "epoch": 0.07842793549009093,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 1538
+    },
+    {
+      "epoch": 0.07847892894619633,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9169,
+      "step": 1539
+    },
+    {
+      "epoch": 0.07852992240230172,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9184,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07858091585840711,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.902,
+      "step": 1541
+    },
+    {
+      "epoch": 0.0786319093145125,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.927,
+      "step": 1542
+    },
+    {
+      "epoch": 0.07868290277061789,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9225,
+      "step": 1543
+    },
+    {
+      "epoch": 0.07873389622672328,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.9213,
+      "step": 1544
+    },
+    {
+      "epoch": 0.07878488968282867,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9201,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07883588313893407,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9136,
+      "step": 1546
+    },
+    {
+      "epoch": 0.07888687659503946,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.932,
+      "step": 1547
+    },
+    {
+      "epoch": 0.07893787005114485,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 1548
+    },
+    {
+      "epoch": 0.07898886350725023,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9205,
+      "step": 1549
+    },
+    {
+      "epoch": 0.07903985696335562,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9343,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07909085041946101,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9116,
+      "step": 1551
+    },
+    {
+      "epoch": 0.0791418438755664,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.902,
+      "step": 1552
+    },
+    {
+      "epoch": 0.0791928373316718,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8979,
+      "step": 1553
+    },
+    {
+      "epoch": 0.07924383078777719,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9343,
+      "step": 1554
+    },
+    {
+      "epoch": 0.07929482424388258,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9284,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07934581769998797,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8977,
+      "step": 1556
+    },
+    {
+      "epoch": 0.07939681115609336,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.888,
+      "step": 1557
+    },
+    {
+      "epoch": 0.07944780461219875,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9279,
+      "step": 1558
+    },
+    {
+      "epoch": 0.07949879806830414,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9474,
+      "step": 1559
+    },
+    {
+      "epoch": 0.07954979152440954,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9413,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07960078498051493,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9083,
+      "step": 1561
+    },
+    {
+      "epoch": 0.07965177843662032,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9061,
+      "step": 1562
+    },
+    {
+      "epoch": 0.0797027718927257,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9282,
+      "step": 1563
+    },
+    {
+      "epoch": 0.0797537653488311,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1564
+    },
+    {
+      "epoch": 0.07980475880493648,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9653,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07985575226104187,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9296,
+      "step": 1566
+    },
+    {
+      "epoch": 0.07990674571714727,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9069,
+      "step": 1567
+    },
+    {
+      "epoch": 0.07995773917325266,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9246,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08000873262935805,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9214,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08005972608546344,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9136,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08011071954156883,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9045,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08016171299767422,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9302,
+      "step": 1572
+    },
+    {
+      "epoch": 0.0802127064537796,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9365,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08026369990988501,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9173,
+      "step": 1574
+    },
+    {
+      "epoch": 0.0803146933659904,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9036,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08036568682209579,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9068,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08041668027820117,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9105,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08046767373430656,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9516,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08051866719041195,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9029,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08056966064651734,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9237,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08062065410262274,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08067164755872813,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9185,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08072264101483352,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9384,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08077363447093891,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8757,
+      "step": 1584
+    },
+    {
+      "epoch": 0.0808246279270443,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9241,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08087562138314969,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9239,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08092661483925508,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9221,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08097760829536048,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9314,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08102860175146587,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9167,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08107959520757126,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9121,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08113058866367664,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9062,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08118158211978203,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9199,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08123257557588742,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9186,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08128356903199281,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9241,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08133456248809821,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 1595
+    },
+    {
+      "epoch": 0.0813855559442036,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9171,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08143654940030899,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9281,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08148754285641438,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.9082,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08153853631251977,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9358,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08158952976862516,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9231,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08164052322473055,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9122,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08169151668083595,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9128,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08174251013694134,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9083,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08179350359304673,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9235,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08184449704915212,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8867,
+      "step": 1605
+    },
+    {
+      "epoch": 0.0818954905052575,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9111,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08194648396136289,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9032,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08199747741746828,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9189,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08204847087357368,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9073,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08209946432967907,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9308,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08215045778578446,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8989,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08220145124188985,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8963,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08225244469799524,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08230343815410063,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 1614
+    },
+    {
+      "epoch": 0.08235443161020602,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9093,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08240542506631142,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.906,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08245641852241681,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8962,
+      "step": 1617
+    },
+    {
+      "epoch": 0.0825074119785222,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.915,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08255840543462759,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9438,
+      "step": 1619
+    },
+    {
+      "epoch": 0.08260939889073297,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9126,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08266039234683836,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9155,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08271138580294375,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9188,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08276237925904915,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9473,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08281337271515454,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08286436617125993,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9292,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08291535962736532,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08296635308347071,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9126,
+      "step": 1627
+    },
+    {
+      "epoch": 0.0830173465395761,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9434,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08306833999568149,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9306,
+      "step": 1629
+    },
+    {
+      "epoch": 0.08311933345178689,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08317032690789228,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08322132036399767,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08327231382010306,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9105,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08332330727620844,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9344,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08337430073231383,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9307,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08342529418841922,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08347628764452462,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8891,
+      "step": 1637
+    },
+    {
+      "epoch": 0.08352728110063001,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9031,
+      "step": 1638
+    },
+    {
+      "epoch": 0.0835782745567354,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9097,
+      "step": 1639
+    },
+    {
+      "epoch": 0.08362926801284079,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9079,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08368026146894618,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9162,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08373125492505157,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9303,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08378224838115696,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9046,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08383324183726236,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9039,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08388423529336775,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9261,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08393522874947314,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9151,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08398622220557853,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8946,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08403721566168391,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9183,
+      "step": 1648
+    },
+    {
+      "epoch": 0.0840882091177893,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8997,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08413920257389469,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9357,
+      "step": 1650
+    },
+    {
+      "epoch": 0.0841901960300001,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08424118948610548,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9051,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08429218294221087,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9007,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08434317639831626,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8954,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08439416985442165,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9148,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08444516331052704,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.9128,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08449615676663243,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9034,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08454715022273783,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9135,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08459814367884322,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9154,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08464913713494861,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 1660
+    },
+    {
+      "epoch": 0.084700130591054,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9337,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08475112404715938,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9204,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08480211750326477,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9162,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08485311095937016,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8863,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08490410441547555,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9049,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08495509787158095,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9127,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08500609132768634,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9236,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08505708478379173,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.924,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08510807823989712,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9211,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08515907169600251,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9097,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0852100651521079,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 1671
+    },
+    {
+      "epoch": 0.08526105860821329,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9028,
+      "step": 1672
+    },
+    {
+      "epoch": 0.08531205206431869,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1673
+    },
+    {
+      "epoch": 0.08536304552042408,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9173,
+      "step": 1674
+    },
+    {
+      "epoch": 0.08541403897652947,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9017,
+      "step": 1675
+    },
+    {
+      "epoch": 0.08546503243263485,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9183,
+      "step": 1676
+    },
+    {
+      "epoch": 0.08551602588874024,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9311,
+      "step": 1677
+    },
+    {
+      "epoch": 0.08556701934484563,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 1678
+    },
+    {
+      "epoch": 0.08561801280095102,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9197,
+      "step": 1679
+    },
+    {
+      "epoch": 0.08566900625705642,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9087,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08571999971316181,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.886,
+      "step": 1681
+    },
+    {
+      "epoch": 0.0857709931692672,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9029,
+      "step": 1682
+    },
+    {
+      "epoch": 0.08582198662537259,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9143,
+      "step": 1683
+    },
+    {
+      "epoch": 0.08587298008147798,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8988,
+      "step": 1684
+    },
+    {
+      "epoch": 0.08592397353758337,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9359,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08597496699368876,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9276,
+      "step": 1686
+    },
+    {
+      "epoch": 0.08602596044979416,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.932,
+      "step": 1687
+    },
+    {
+      "epoch": 0.08607695390589955,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8968,
+      "step": 1688
+    },
+    {
+      "epoch": 0.08612794736200494,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9206,
+      "step": 1689
+    },
+    {
+      "epoch": 0.08617894081811033,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9179,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08622993427421571,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9151,
+      "step": 1691
+    },
+    {
+      "epoch": 0.0862809277303211,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9071,
+      "step": 1692
+    },
+    {
+      "epoch": 0.08633192118642649,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9017,
+      "step": 1693
+    },
+    {
+      "epoch": 0.0863829146425319,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9248,
+      "step": 1694
+    },
+    {
+      "epoch": 0.08643390809863728,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08648490155474267,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9349,
+      "step": 1696
+    },
+    {
+      "epoch": 0.08653589501084806,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8986,
+      "step": 1697
+    },
+    {
+      "epoch": 0.08658688846695345,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9423,
+      "step": 1698
+    },
+    {
+      "epoch": 0.08663788192305884,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9373,
+      "step": 1699
+    },
+    {
+      "epoch": 0.08668887537916423,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9333,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08673986883526963,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9156,
+      "step": 1701
+    },
+    {
+      "epoch": 0.08679086229137502,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9036,
+      "step": 1702
+    },
+    {
+      "epoch": 0.0868418557474804,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9159,
+      "step": 1703
+    },
+    {
+      "epoch": 0.0868928492035858,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9229,
+      "step": 1704
+    },
+    {
+      "epoch": 0.08694384265969118,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9251,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08699483611579657,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.9022,
+      "step": 1706
+    },
+    {
+      "epoch": 0.08704582957190196,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9393,
+      "step": 1707
+    },
+    {
+      "epoch": 0.08709682302800736,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9199,
+      "step": 1708
+    },
+    {
+      "epoch": 0.08714781648411275,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9229,
+      "step": 1709
+    },
+    {
+      "epoch": 0.08719880994021814,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.889,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08724980339632353,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.912,
+      "step": 1711
+    },
+    {
+      "epoch": 0.08730079685242892,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9428,
+      "step": 1712
+    },
+    {
+      "epoch": 0.08735179030853431,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8993,
+      "step": 1713
+    },
+    {
+      "epoch": 0.0874027837646397,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8994,
+      "step": 1714
+    },
+    {
+      "epoch": 0.0874537772207451,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9132,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08750477067685049,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.897,
+      "step": 1716
+    },
+    {
+      "epoch": 0.08755576413295588,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9304,
+      "step": 1717
+    },
+    {
+      "epoch": 0.08760675758906127,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9131,
+      "step": 1718
+    },
+    {
+      "epoch": 0.08765775104516665,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.92,
+      "step": 1719
+    },
+    {
+      "epoch": 0.08770874450127204,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9058,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08775973795737743,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8848,
+      "step": 1721
+    },
+    {
+      "epoch": 0.08781073141348283,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8912,
+      "step": 1722
+    },
+    {
+      "epoch": 0.08786172486958822,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.911,
+      "step": 1723
+    },
+    {
+      "epoch": 0.08791271832569361,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1724
+    },
+    {
+      "epoch": 0.087963711781799,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9133,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08801470523790439,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9047,
+      "step": 1726
+    },
+    {
+      "epoch": 0.08806569869400978,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9046,
+      "step": 1727
+    },
+    {
+      "epoch": 0.08811669215011517,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9086,
+      "step": 1728
+    },
+    {
+      "epoch": 0.08816768560622057,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 1729
+    },
+    {
+      "epoch": 0.08821867906232596,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9296,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08826967251843135,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8867,
+      "step": 1731
+    },
+    {
+      "epoch": 0.08832066597453674,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8982,
+      "step": 1732
+    },
+    {
+      "epoch": 0.08837165943064212,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 1733
+    },
+    {
+      "epoch": 0.08842265288674751,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9095,
+      "step": 1734
+    },
+    {
+      "epoch": 0.0884736463428529,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.9106,
+      "step": 1735
+    },
+    {
+      "epoch": 0.0885246397989583,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.894,
+      "step": 1736
+    },
+    {
+      "epoch": 0.0885756332550637,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9247,
+      "step": 1737
+    },
+    {
+      "epoch": 0.08862662671116908,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8967,
+      "step": 1738
+    },
+    {
+      "epoch": 0.08867762016727447,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9044,
+      "step": 1739
+    },
+    {
+      "epoch": 0.08872861362337986,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9138,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08877960707948525,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9032,
+      "step": 1741
+    },
+    {
+      "epoch": 0.08883060053559064,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 1742
+    },
+    {
+      "epoch": 0.08888159399169604,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9261,
+      "step": 1743
+    },
+    {
+      "epoch": 0.08893258744780143,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.9061,
+      "step": 1744
+    },
+    {
+      "epoch": 0.08898358090390682,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9059,
+      "step": 1745
+    },
+    {
+      "epoch": 0.0890345743600122,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9224,
+      "step": 1746
+    },
+    {
+      "epoch": 0.0890855678161176,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 1747
+    },
+    {
+      "epoch": 0.08913656127222298,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 1748
+    },
+    {
+      "epoch": 0.08918755472832837,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.915,
+      "step": 1749
+    },
+    {
+      "epoch": 0.08923854818443377,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9214,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08928954164053916,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9013,
+      "step": 1751
+    },
+    {
+      "epoch": 0.08934053509664455,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9188,
+      "step": 1752
+    },
+    {
+      "epoch": 0.08939152855274994,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9071,
+      "step": 1753
+    },
+    {
+      "epoch": 0.08944252200885533,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9131,
+      "step": 1754
+    },
+    {
+      "epoch": 0.08949351546496072,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8993,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08954450892106611,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9041,
+      "step": 1756
+    },
+    {
+      "epoch": 0.08959550237717151,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8886,
+      "step": 1757
+    },
+    {
+      "epoch": 0.0896464958332769,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9054,
+      "step": 1758
+    },
+    {
+      "epoch": 0.08969748928938229,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8898,
+      "step": 1759
+    },
+    {
+      "epoch": 0.08974848274548768,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9172,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08979947620159306,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9115,
+      "step": 1761
+    },
+    {
+      "epoch": 0.08985046965769845,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9096,
+      "step": 1762
+    },
+    {
+      "epoch": 0.08990146311380384,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8985,
+      "step": 1763
+    },
+    {
+      "epoch": 0.08995245656990924,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09000345002601463,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.922,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09005444348212002,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9229,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09010543693822541,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8992,
+      "step": 1767
+    },
+    {
+      "epoch": 0.0901564303943308,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9233,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09020742385043619,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9242,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09025841730654158,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09030941076264698,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.912,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09036040421875237,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9248,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09041139767485776,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9067,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09046239113096315,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9041,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09051338458706853,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9016,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09056437804317392,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09061537149927931,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9223,
+      "step": 1777
+    },
+    {
+      "epoch": 0.09066636495538472,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9215,
+      "step": 1778
+    },
+    {
+      "epoch": 0.0907173584114901,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09076835186759549,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8914,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09081934532370088,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9267,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09087033877980627,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.9464,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09092133223591166,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8928,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09097232569201705,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9129,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09102331914812245,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9058,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09107431260422784,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09112530606033323,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8948,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09117629951643862,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9253,
+      "step": 1788
+    },
+    {
+      "epoch": 0.091227292972544,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9136,
+      "step": 1789
+    },
+    {
+      "epoch": 0.0912782864286494,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9151,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09132927988475478,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9093,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09138027334086019,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9197,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09143126679696557,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8995,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09148226025307096,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9024,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09153325370917635,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09158424716528174,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9104,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09163524062138713,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09168623407749252,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09173722753359792,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09178822098970331,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9204,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0918392144458087,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9146,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09189020790191409,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9124,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09194120135801948,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9236,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09199219481412486,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8969,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09204318827023025,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8977,
+      "step": 1805
+    },
+    {
+      "epoch": 0.09209418172633566,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.906,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09214517518244104,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9201,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09219616863854643,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9008,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09224716209465182,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.917,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09229815555075721,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8923,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0923491490068626,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9237,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09240014246296799,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9026,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09245113591907339,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9113,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09250212937517878,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9009,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09255312283128417,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1815
+    },
+    {
+      "epoch": 0.09260411628738956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9113,
+      "step": 1816
+    },
+    {
+      "epoch": 0.09265510974349495,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9153,
+      "step": 1817
+    },
+    {
+      "epoch": 0.09270610319960033,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9285,
+      "step": 1818
+    },
+    {
+      "epoch": 0.09275709665570572,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9079,
+      "step": 1819
+    },
+    {
+      "epoch": 0.09280809011181113,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9327,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09285908356791651,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9026,
+      "step": 1821
+    },
+    {
+      "epoch": 0.0929100770240219,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.882,
+      "step": 1822
+    },
+    {
+      "epoch": 0.09296107048012729,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9169,
+      "step": 1823
+    },
+    {
+      "epoch": 0.09301206393623268,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9291,
+      "step": 1824
+    },
+    {
+      "epoch": 0.09306305739233807,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9186,
+      "step": 1825
+    },
+    {
+      "epoch": 0.09311405084844346,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9293,
+      "step": 1826
+    },
+    {
+      "epoch": 0.09316504430454886,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9378,
+      "step": 1827
+    },
+    {
+      "epoch": 0.09321603776065425,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9047,
+      "step": 1828
+    },
+    {
+      "epoch": 0.09326703121675964,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.906,
+      "step": 1829
+    },
+    {
+      "epoch": 0.09331802467286503,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8952,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09336901812897042,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9073,
+      "step": 1831
+    },
+    {
+      "epoch": 0.0934200115850758,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9083,
+      "step": 1832
+    },
+    {
+      "epoch": 0.0934710050411812,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 1833
+    },
+    {
+      "epoch": 0.0935219984972866,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.9112,
+      "step": 1834
+    },
+    {
+      "epoch": 0.09357299195339198,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9295,
+      "step": 1835
+    },
+    {
+      "epoch": 0.09362398540949737,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8845,
+      "step": 1836
+    },
+    {
+      "epoch": 0.09367497886560276,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9002,
+      "step": 1837
+    },
+    {
+      "epoch": 0.09372597232170815,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9185,
+      "step": 1838
+    },
+    {
+      "epoch": 0.09377696577781354,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9189,
+      "step": 1839
+    },
+    {
+      "epoch": 0.09382795923391893,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09387895269002433,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8839,
+      "step": 1841
+    },
+    {
+      "epoch": 0.09392994614612972,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 1842
+    },
+    {
+      "epoch": 0.09398093960223511,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 1843
+    },
+    {
+      "epoch": 0.0940319330583405,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9189,
+      "step": 1844
+    },
+    {
+      "epoch": 0.09408292651444589,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 1845
+    },
+    {
+      "epoch": 0.09413391997055127,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8769,
+      "step": 1846
+    },
+    {
+      "epoch": 0.09418491342665666,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 1847
+    },
+    {
+      "epoch": 0.09423590688276207,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9148,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09428690033886745,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9369,
+      "step": 1849
+    },
+    {
+      "epoch": 0.09433789379497284,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8899,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09438888725107823,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1851
+    },
+    {
+      "epoch": 0.09443988070718362,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8974,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09449087416328901,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9105,
+      "step": 1853
+    },
+    {
+      "epoch": 0.0945418676193944,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.885,
+      "step": 1854
+    },
+    {
+      "epoch": 0.0945928610754998,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9234,
+      "step": 1855
+    },
+    {
+      "epoch": 0.09464385453160519,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09469484798771058,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8716,
+      "step": 1857
+    },
+    {
+      "epoch": 0.09474584144381597,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9037,
+      "step": 1858
+    },
+    {
+      "epoch": 0.09479683489992136,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9336,
+      "step": 1859
+    },
+    {
+      "epoch": 0.09484782835602674,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09489882181213213,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09494981526823754,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9239,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09500080872434293,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8995,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09505180218044831,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9021,
+      "step": 1864
+    },
+    {
+      "epoch": 0.0951027956365537,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9049,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09515378909265909,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.9066,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09520478254876448,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9142,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09525577600486987,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8931,
+      "step": 1868
+    },
+    {
+      "epoch": 0.09530676946097527,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1869
+    },
+    {
+      "epoch": 0.09535776291708066,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8845,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09540875637318605,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9275,
+      "step": 1871
+    },
+    {
+      "epoch": 0.09545974982929144,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 1872
+    },
+    {
+      "epoch": 0.09551074328539683,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8991,
+      "step": 1873
+    },
+    {
+      "epoch": 0.09556173674150222,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 1874
+    },
+    {
+      "epoch": 0.0956127301976076,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9181,
+      "step": 1875
+    },
+    {
+      "epoch": 0.095663723653713,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9225,
+      "step": 1876
+    },
+    {
+      "epoch": 0.0957147171098184,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 1877
+    },
+    {
+      "epoch": 0.09576571056592378,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8986,
+      "step": 1878
+    },
+    {
+      "epoch": 0.09581670402202917,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9008,
+      "step": 1879
+    },
+    {
+      "epoch": 0.09586769747813456,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8827,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09591869093423995,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1881
+    },
+    {
+      "epoch": 0.09596968439034534,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 1882
+    },
+    {
+      "epoch": 0.09602067784645074,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.9053,
+      "step": 1883
+    },
+    {
+      "epoch": 0.09607167130255613,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9236,
+      "step": 1884
+    },
+    {
+      "epoch": 0.09612266475866152,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9276,
+      "step": 1885
+    },
+    {
+      "epoch": 0.09617365821476691,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.923,
+      "step": 1886
+    },
+    {
+      "epoch": 0.0962246516708723,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9243,
+      "step": 1887
+    },
+    {
+      "epoch": 0.09627564512697769,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 1888
+    },
+    {
+      "epoch": 0.09632663858308307,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 1889
+    },
+    {
+      "epoch": 0.09637763203918848,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9241,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09642862549529387,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 1891
+    },
+    {
+      "epoch": 0.09647961895139925,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9203,
+      "step": 1892
+    },
+    {
+      "epoch": 0.09653061240750464,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9102,
+      "step": 1893
+    },
+    {
+      "epoch": 0.09658160586361003,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09663259931971542,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8883,
+      "step": 1895
+    },
+    {
+      "epoch": 0.09668359277582081,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9196,
+      "step": 1896
+    },
+    {
+      "epoch": 0.09673458623192621,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 1897
+    },
+    {
+      "epoch": 0.0967855796880316,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9143,
+      "step": 1898
+    },
+    {
+      "epoch": 0.09683657314413699,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9209,
+      "step": 1899
+    },
+    {
+      "epoch": 0.09688756660024238,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.9192,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09693856005634777,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8825,
+      "step": 1901
+    },
+    {
+      "epoch": 0.09698955351245316,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 1902
+    },
+    {
+      "epoch": 0.09704054696855854,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 1903
+    },
+    {
+      "epoch": 0.09709154042466395,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.902,
+      "step": 1904
+    },
+    {
+      "epoch": 0.09714253388076934,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 1905
+    },
+    {
+      "epoch": 0.09719352733687472,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 1906
+    },
+    {
+      "epoch": 0.09724452079298011,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8957,
+      "step": 1907
+    },
+    {
+      "epoch": 0.0972955142490855,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9124,
+      "step": 1908
+    },
+    {
+      "epoch": 0.09734650770519089,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.9097,
+      "step": 1909
+    },
+    {
+      "epoch": 0.09739750116129628,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9212,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09744849461740168,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9015,
+      "step": 1911
+    },
+    {
+      "epoch": 0.09749948807350707,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.912,
+      "step": 1912
+    },
+    {
+      "epoch": 0.09755048152961246,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.912,
+      "step": 1913
+    },
+    {
+      "epoch": 0.09760147498571785,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.911,
+      "step": 1914
+    },
+    {
+      "epoch": 0.09765246844182324,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 1915
+    },
+    {
+      "epoch": 0.09770346189792863,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9031,
+      "step": 1916
+    },
+    {
+      "epoch": 0.09775445535403401,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 1917
+    },
+    {
+      "epoch": 0.09780544881013942,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9054,
+      "step": 1918
+    },
+    {
+      "epoch": 0.0978564422662448,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 1919
+    },
+    {
+      "epoch": 0.0979074357223502,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9343,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09795842917845558,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.903,
+      "step": 1921
+    },
+    {
+      "epoch": 0.09800942263456097,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 1922
+    },
+    {
+      "epoch": 0.09806041609066636,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9279,
+      "step": 1923
+    },
+    {
+      "epoch": 0.09811140954677175,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9045,
+      "step": 1924
+    },
+    {
+      "epoch": 0.09816240300287715,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1925
+    },
+    {
+      "epoch": 0.09821339645898254,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8987,
+      "step": 1926
+    },
+    {
+      "epoch": 0.09826438991508793,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 1927
+    },
+    {
+      "epoch": 0.09831538337119332,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8681,
+      "step": 1928
+    },
+    {
+      "epoch": 0.09836637682729871,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9028,
+      "step": 1929
+    },
+    {
+      "epoch": 0.0984173702834041,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9187,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09846836373950948,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 1931
+    },
+    {
+      "epoch": 0.09851935719561489,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8883,
+      "step": 1932
+    },
+    {
+      "epoch": 0.09857035065172028,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 1933
+    },
+    {
+      "epoch": 0.09862134410782566,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9038,
+      "step": 1934
+    },
+    {
+      "epoch": 0.09867233756393105,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8991,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09872333102003644,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 1936
+    },
+    {
+      "epoch": 0.09877432447614183,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9265,
+      "step": 1937
+    },
+    {
+      "epoch": 0.09882531793224722,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9247,
+      "step": 1938
+    },
+    {
+      "epoch": 0.09887631138835262,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8773,
+      "step": 1939
+    },
+    {
+      "epoch": 0.09892730484445801,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9238,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0989782983005634,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9065,
+      "step": 1941
+    },
+    {
+      "epoch": 0.09902929175666879,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.9241,
+      "step": 1942
+    },
+    {
+      "epoch": 0.09908028521277418,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9097,
+      "step": 1943
+    },
+    {
+      "epoch": 0.09913127866887957,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8939,
+      "step": 1944
+    },
+    {
+      "epoch": 0.09918227212498495,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9082,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09923326558109036,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 1946
+    },
+    {
+      "epoch": 0.09928425903719575,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 1947
+    },
+    {
+      "epoch": 0.09933525249330113,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.897,
+      "step": 1948
+    },
+    {
+      "epoch": 0.09938624594940652,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.9071,
+      "step": 1949
+    },
+    {
+      "epoch": 0.09943723940551191,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8859,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0994882328616173,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8846,
+      "step": 1951
+    },
+    {
+      "epoch": 0.09953922631772269,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.9317,
+      "step": 1952
+    },
+    {
+      "epoch": 0.09959021977382809,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.9233,
+      "step": 1953
+    },
+    {
+      "epoch": 0.09964121322993348,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.9121,
+      "step": 1954
+    },
+    {
+      "epoch": 0.09969220668603887,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09974320014214426,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8966,
+      "step": 1956
+    },
+    {
+      "epoch": 0.09979419359824965,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9131,
+      "step": 1957
+    },
+    {
+      "epoch": 0.09984518705435504,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8923,
+      "step": 1958
+    },
+    {
+      "epoch": 0.09989618051046043,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 1959
+    },
+    {
+      "epoch": 0.09994717396656583,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9112,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09999816742267122,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9207,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1000491608787766,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9129,
+      "step": 1962
+    },
+    {
+      "epoch": 0.100100154334882,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9079,
+      "step": 1963
+    },
+    {
+      "epoch": 0.10015114779098738,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.913,
+      "step": 1964
+    },
+    {
+      "epoch": 0.10020214124709277,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9153,
+      "step": 1965
+    },
+    {
+      "epoch": 0.10025313470319816,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.898,
+      "step": 1966
+    },
+    {
+      "epoch": 0.10030412815930356,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.927,
+      "step": 1967
+    },
+    {
+      "epoch": 0.10035512161540895,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 1968
+    },
+    {
+      "epoch": 0.10040611507151434,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9182,
+      "step": 1969
+    },
+    {
+      "epoch": 0.10045710852761973,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9165,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10050810198372512,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1005590954398305,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.925,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1006100888959359,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8968,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1006610823520413,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8967,
+      "step": 1974
+    },
+    {
+      "epoch": 0.10071207580814669,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9012,
+      "step": 1975
+    },
+    {
+      "epoch": 0.10076306926425208,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.909,
+      "step": 1976
+    },
+    {
+      "epoch": 0.10081406272035746,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9061,
+      "step": 1977
+    },
+    {
+      "epoch": 0.10086505617646285,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.907,
+      "step": 1978
+    },
+    {
+      "epoch": 0.10091604963256824,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9226,
+      "step": 1979
+    },
+    {
+      "epoch": 0.10096704308867363,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9115,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10101803654477903,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 1981
+    },
+    {
+      "epoch": 0.10106903000088442,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9156,
+      "step": 1982
+    },
+    {
+      "epoch": 0.10112002345698981,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8813,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1011710169130952,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9203,
+      "step": 1984
+    },
+    {
+      "epoch": 0.10122201036920059,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 1985
+    },
+    {
+      "epoch": 0.10127300382530598,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8952,
+      "step": 1986
+    },
+    {
+      "epoch": 0.10132399728141137,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9134,
+      "step": 1987
+    },
+    {
+      "epoch": 0.10137499073751677,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 1988
+    },
+    {
+      "epoch": 0.10142598419362216,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.9218,
+      "step": 1989
+    },
+    {
+      "epoch": 0.10147697764972755,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9025,
+      "step": 1990
+    },
+    {
+      "epoch": 0.10152797110583293,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9073,
+      "step": 1991
+    },
+    {
+      "epoch": 0.10157896456193832,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8729,
+      "step": 1992
+    },
+    {
+      "epoch": 0.10162995801804371,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8818,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1016809514741491,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8761,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1017319449302545,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9136,
+      "step": 1995
+    },
+    {
+      "epoch": 0.10178293838635989,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9023,
+      "step": 1996
+    },
+    {
+      "epoch": 0.10183393184246528,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8993,
+      "step": 1997
+    },
+    {
+      "epoch": 0.10188492529857067,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 1998
+    },
+    {
+      "epoch": 0.10193591875467606,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9056,
+      "step": 1999
+    },
+    {
+      "epoch": 0.10198691221078145,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8957,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10203790566688684,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 2001
+    },
+    {
+      "epoch": 0.10208889912299224,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.942,
+      "step": 2002
+    },
+    {
+      "epoch": 0.10213989257909763,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 2003
+    },
+    {
+      "epoch": 0.10219088603520302,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1022418794913084,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8838,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1022928729474138,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.899,
+      "step": 2006
+    },
+    {
+      "epoch": 0.10234386640351918,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9077,
+      "step": 2007
+    },
+    {
+      "epoch": 0.10239485985962457,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.897,
+      "step": 2008
+    },
+    {
+      "epoch": 0.10244585331572997,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.9148,
+      "step": 2009
+    },
+    {
+      "epoch": 0.10249684677183536,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 2010
+    },
+    {
+      "epoch": 0.10254784022794075,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.896,
+      "step": 2011
+    },
+    {
+      "epoch": 0.10259883368404614,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 2012
+    },
+    {
+      "epoch": 0.10264982714015153,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.9015,
+      "step": 2013
+    },
+    {
+      "epoch": 0.10270082059625692,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8906,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1027518140523623,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2015
+    },
+    {
+      "epoch": 0.10280280750846771,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9319,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1028538009645731,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 2017
+    },
+    {
+      "epoch": 0.10290479442067849,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9034,
+      "step": 2018
+    },
+    {
+      "epoch": 0.10295578787678387,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9139,
+      "step": 2019
+    },
+    {
+      "epoch": 0.10300678133288926,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9359,
+      "step": 2020
+    },
+    {
+      "epoch": 0.10305777478899465,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8971,
+      "step": 2021
+    },
+    {
+      "epoch": 0.10310876824510004,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.9155,
+      "step": 2022
+    },
+    {
+      "epoch": 0.10315976170120544,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 2023
+    },
+    {
+      "epoch": 0.10321075515731083,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.932,
+      "step": 2024
+    },
+    {
+      "epoch": 0.10326174861341622,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9131,
+      "step": 2025
+    },
+    {
+      "epoch": 0.10331274206952161,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 2026
+    },
+    {
+      "epoch": 0.103363735525627,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 2027
+    },
+    {
+      "epoch": 0.10341472898173239,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8943,
+      "step": 2028
+    },
+    {
+      "epoch": 0.10346572243783778,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9336,
+      "step": 2029
+    },
+    {
+      "epoch": 0.10351671589394318,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10356770935004857,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.9195,
+      "step": 2031
+    },
+    {
+      "epoch": 0.10361870280615396,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9138,
+      "step": 2032
+    },
+    {
+      "epoch": 0.10366969626225934,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9255,
+      "step": 2033
+    },
+    {
+      "epoch": 0.10372068971836473,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8972,
+      "step": 2034
+    },
+    {
+      "epoch": 0.10377168317447012,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.9082,
+      "step": 2035
+    },
+    {
+      "epoch": 0.10382267663057551,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9071,
+      "step": 2036
+    },
+    {
+      "epoch": 0.10387367008668091,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1039246635427863,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8962,
+      "step": 2038
+    },
+    {
+      "epoch": 0.10397565699889169,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9057,
+      "step": 2039
+    },
+    {
+      "epoch": 0.10402665045499708,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8888,
+      "step": 2040
+    },
+    {
+      "epoch": 0.10407764391110247,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8973,
+      "step": 2041
+    },
+    {
+      "epoch": 0.10412863736720786,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9062,
+      "step": 2042
+    },
+    {
+      "epoch": 0.10417963082331325,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9025,
+      "step": 2043
+    },
+    {
+      "epoch": 0.10423062427941865,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 2044
+    },
+    {
+      "epoch": 0.10428161773552404,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 2045
+    },
+    {
+      "epoch": 0.10433261119162943,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 2046
+    },
+    {
+      "epoch": 0.10438360464773482,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8964,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1044345981038402,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.9032,
+      "step": 2048
+    },
+    {
+      "epoch": 0.10448559155994559,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.9055,
+      "step": 2049
+    },
+    {
+      "epoch": 0.10453658501605098,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8984,
+      "step": 2050
+    },
+    {
+      "epoch": 0.10458757847215638,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.9014,
+      "step": 2051
+    },
+    {
+      "epoch": 0.10463857192826177,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8588,
+      "step": 2052
+    },
+    {
+      "epoch": 0.10468956538436716,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.918,
+      "step": 2053
+    },
+    {
+      "epoch": 0.10474055884047255,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8999,
+      "step": 2054
+    },
+    {
+      "epoch": 0.10479155229657794,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 2055
+    },
+    {
+      "epoch": 0.10484254575268333,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.9024,
+      "step": 2056
+    },
+    {
+      "epoch": 0.10489353920878872,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8912,
+      "step": 2057
+    },
+    {
+      "epoch": 0.10494453266489412,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8878,
+      "step": 2058
+    },
+    {
+      "epoch": 0.10499552612099951,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9034,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1050465195771049,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.893,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10509751303321029,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8976,
+      "step": 2061
+    },
+    {
+      "epoch": 0.10514850648931567,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9193,
+      "step": 2062
+    },
+    {
+      "epoch": 0.10519949994542106,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9124,
+      "step": 2063
+    },
+    {
+      "epoch": 0.10525049340152645,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8934,
+      "step": 2064
+    },
+    {
+      "epoch": 0.10530148685763185,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8841,
+      "step": 2065
+    },
+    {
+      "epoch": 0.10535248031373724,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8976,
+      "step": 2066
+    },
+    {
+      "epoch": 0.10540347376984263,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8989,
+      "step": 2067
+    },
+    {
+      "epoch": 0.10545446722594802,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8748,
+      "step": 2068
+    },
+    {
+      "epoch": 0.10550546068205341,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1055564541381588,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10560744759426419,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9021,
+      "step": 2071
+    },
+    {
+      "epoch": 0.10565844105036959,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9208,
+      "step": 2072
+    },
+    {
+      "epoch": 0.10570943450647498,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8919,
+      "step": 2073
+    },
+    {
+      "epoch": 0.10576042796258037,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9013,
+      "step": 2074
+    },
+    {
+      "epoch": 0.10581142141868576,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.877,
+      "step": 2075
+    },
+    {
+      "epoch": 0.10586241487479114,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 2076
+    },
+    {
+      "epoch": 0.10591340833089653,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9153,
+      "step": 2077
+    },
+    {
+      "epoch": 0.10596440178700192,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8778,
+      "step": 2078
+    },
+    {
+      "epoch": 0.10601539524310732,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.929,
+      "step": 2079
+    },
+    {
+      "epoch": 0.10606638869921271,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1061173821553181,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8966,
+      "step": 2081
+    },
+    {
+      "epoch": 0.10616837561142349,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 2082
+    },
+    {
+      "epoch": 0.10621936906752888,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 2083
+    },
+    {
+      "epoch": 0.10627036252363427,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8964,
+      "step": 2084
+    },
+    {
+      "epoch": 0.10632135597973966,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 2085
+    },
+    {
+      "epoch": 0.10637234943584506,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9135,
+      "step": 2086
+    },
+    {
+      "epoch": 0.10642334289195045,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 2087
+    },
+    {
+      "epoch": 0.10647433634805584,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8943,
+      "step": 2088
+    },
+    {
+      "epoch": 0.10652532980416123,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8728,
+      "step": 2089
+    },
+    {
+      "epoch": 0.10657632326026661,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 2090
+    },
+    {
+      "epoch": 0.106627316716372,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 2091
+    },
+    {
+      "epoch": 0.10667831017247739,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1067293036285828,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9047,
+      "step": 2093
+    },
+    {
+      "epoch": 0.10678029708468818,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 2094
+    },
+    {
+      "epoch": 0.10683129054079357,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 2095
+    },
+    {
+      "epoch": 0.10688228399689896,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9362,
+      "step": 2096
+    },
+    {
+      "epoch": 0.10693327745300435,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 2097
+    },
+    {
+      "epoch": 0.10698427090910974,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9111,
+      "step": 2098
+    },
+    {
+      "epoch": 0.10703526436521513,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.893,
+      "step": 2099
+    },
+    {
+      "epoch": 0.10708625782132053,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9053,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10713725127742592,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.9096,
+      "step": 2101
+    },
+    {
+      "epoch": 0.10718824473353131,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1072392381896367,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9049,
+      "step": 2103
+    },
+    {
+      "epoch": 0.10729023164574208,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8839,
+      "step": 2104
+    },
+    {
+      "epoch": 0.10734122510184747,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8951,
+      "step": 2105
+    },
+    {
+      "epoch": 0.10739221855795286,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8943,
+      "step": 2106
+    },
+    {
+      "epoch": 0.10744321201405826,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8816,
+      "step": 2107
+    },
+    {
+      "epoch": 0.10749420547016365,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8884,
+      "step": 2108
+    },
+    {
+      "epoch": 0.10754519892626904,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9163,
+      "step": 2109
+    },
+    {
+      "epoch": 0.10759619238237443,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10764718583847982,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8701,
+      "step": 2111
+    },
+    {
+      "epoch": 0.10769817929458521,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9047,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1077491727506906,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 2113
+    },
+    {
+      "epoch": 0.107800166206796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9022,
+      "step": 2114
+    },
+    {
+      "epoch": 0.10785115966290139,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 2115
+    },
+    {
+      "epoch": 0.10790215311900678,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8973,
+      "step": 2116
+    },
+    {
+      "epoch": 0.10795314657511217,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 2117
+    },
+    {
+      "epoch": 0.10800414003121755,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 2118
+    },
+    {
+      "epoch": 0.10805513348732294,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 2119
+    },
+    {
+      "epoch": 0.10810612694342833,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9072,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10815712039953373,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8947,
+      "step": 2121
+    },
+    {
+      "epoch": 0.10820811385563912,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9038,
+      "step": 2122
+    },
+    {
+      "epoch": 0.10825910731174451,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9105,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1083101007678499,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 2124
+    },
+    {
+      "epoch": 0.10836109422395529,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 2125
+    },
+    {
+      "epoch": 0.10841208768006068,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.944,
+      "step": 2126
+    },
+    {
+      "epoch": 0.10846308113616607,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 2127
+    },
+    {
+      "epoch": 0.10851407459227147,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9101,
+      "step": 2128
+    },
+    {
+      "epoch": 0.10856506804837686,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 2129
+    },
+    {
+      "epoch": 0.10861606150448225,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8897,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10866705496058764,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9023,
+      "step": 2131
+    },
+    {
+      "epoch": 0.10871804841669302,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2132
+    },
+    {
+      "epoch": 0.10876904187279841,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8942,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1088200353289038,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 2134
+    },
+    {
+      "epoch": 0.1088710287850092,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8955,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1089220222411146,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8745,
+      "step": 2136
+    },
+    {
+      "epoch": 0.10897301569721998,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 2137
+    },
+    {
+      "epoch": 0.10902400915332537,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.9134,
+      "step": 2138
+    },
+    {
+      "epoch": 0.10907500260943076,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9064,
+      "step": 2139
+    },
+    {
+      "epoch": 0.10912599606553615,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.914,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10917698952164154,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9279,
+      "step": 2141
+    },
+    {
+      "epoch": 0.10922798297774694,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 2142
+    },
+    {
+      "epoch": 0.10927897643385233,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8848,
+      "step": 2143
+    },
+    {
+      "epoch": 0.10932996988995772,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9128,
+      "step": 2144
+    },
+    {
+      "epoch": 0.1093809633460631,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8898,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1094319568021685,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 2146
+    },
+    {
+      "epoch": 0.10948295025827388,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8899,
+      "step": 2147
+    },
+    {
+      "epoch": 0.10953394371437927,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2148
+    },
+    {
+      "epoch": 0.10958493717048468,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 2149
+    },
+    {
+      "epoch": 0.10963593062659006,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.9118,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10968692408269545,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8859,
+      "step": 2151
+    },
+    {
+      "epoch": 0.10973791753880084,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 2152
+    },
+    {
+      "epoch": 0.10978891099490623,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8854,
+      "step": 2153
+    },
+    {
+      "epoch": 0.10983990445101162,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 2154
+    },
+    {
+      "epoch": 0.10989089790711701,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9048,
+      "step": 2155
+    },
+    {
+      "epoch": 0.10994189136322241,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9065,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1099928848193278,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 2157
+    },
+    {
+      "epoch": 0.11004387827543319,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8974,
+      "step": 2158
+    },
+    {
+      "epoch": 0.11009487173153858,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 2159
+    },
+    {
+      "epoch": 0.11014586518764397,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9055,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11019685864374935,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8774,
+      "step": 2161
+    },
+    {
+      "epoch": 0.11024785209985474,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9117,
+      "step": 2162
+    },
+    {
+      "epoch": 0.11029884555596015,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 2163
+    },
+    {
+      "epoch": 0.11034983901206553,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 2164
+    },
+    {
+      "epoch": 0.11040083246817092,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11045182592427631,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9002,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1105028193803817,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8886,
+      "step": 2167
+    },
+    {
+      "epoch": 0.11055381283648709,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8836,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11060480629259248,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9168,
+      "step": 2169
+    },
+    {
+      "epoch": 0.11065579974869788,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8883,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11070679320480327,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 2171
+    },
+    {
+      "epoch": 0.11075778666090866,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.893,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11080878011701405,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 2173
+    },
+    {
+      "epoch": 0.11085977357311944,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9135,
+      "step": 2174
+    },
+    {
+      "epoch": 0.11091076702922482,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9041,
+      "step": 2175
+    },
+    {
+      "epoch": 0.11096176048533021,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9113,
+      "step": 2176
+    },
+    {
+      "epoch": 0.11101275394143562,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8818,
+      "step": 2177
+    },
+    {
+      "epoch": 0.111063747397541,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9133,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1111147408536464,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8801,
+      "step": 2179
+    },
+    {
+      "epoch": 0.11116573430975178,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11121672776585717,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.907,
+      "step": 2181
+    },
+    {
+      "epoch": 0.11126772122196256,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9192,
+      "step": 2182
+    },
+    {
+      "epoch": 0.11131871467806795,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8866,
+      "step": 2183
+    },
+    {
+      "epoch": 0.11136970813417335,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 2184
+    },
+    {
+      "epoch": 0.11142070159027874,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 2185
+    },
+    {
+      "epoch": 0.11147169504638413,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9053,
+      "step": 2186
+    },
+    {
+      "epoch": 0.11152268850248952,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8794,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1115736819585949,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1116246754147003,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.896,
+      "step": 2189
+    },
+    {
+      "epoch": 0.11167566887080568,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8763,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11172666232691109,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8866,
+      "step": 2191
+    },
+    {
+      "epoch": 0.11177765578301647,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8759,
+      "step": 2192
+    },
+    {
+      "epoch": 0.11182864923912186,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9033,
+      "step": 2193
+    },
+    {
+      "epoch": 0.11187964269522725,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8623,
+      "step": 2194
+    },
+    {
+      "epoch": 0.11193063615133264,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8927,
+      "step": 2195
+    },
+    {
+      "epoch": 0.11198162960743803,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 2196
+    },
+    {
+      "epoch": 0.11203262306354342,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.901,
+      "step": 2197
+    },
+    {
+      "epoch": 0.11208361651964882,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9039,
+      "step": 2198
+    },
+    {
+      "epoch": 0.11213460997575421,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8787,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1121856034318596,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11223659688796499,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9013,
+      "step": 2201
+    },
+    {
+      "epoch": 0.11228759034407038,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 2202
+    },
+    {
+      "epoch": 0.11233858380017576,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8739,
+      "step": 2203
+    },
+    {
+      "epoch": 0.11238957725628115,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9019,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11244057071238656,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9176,
+      "step": 2205
+    },
+    {
+      "epoch": 0.11249156416849194,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9047,
+      "step": 2206
+    },
+    {
+      "epoch": 0.11254255762459733,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9061,
+      "step": 2207
+    },
+    {
+      "epoch": 0.11259355108070272,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.9031,
+      "step": 2208
+    },
+    {
+      "epoch": 0.11264454453680811,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.9379,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1126955379929135,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11274653144901889,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 2211
+    },
+    {
+      "epoch": 0.11279752490512429,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9029,
+      "step": 2212
+    },
+    {
+      "epoch": 0.11284851836122968,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 2213
+    },
+    {
+      "epoch": 0.11289951181733507,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9075,
+      "step": 2214
+    },
+    {
+      "epoch": 0.11295050527344046,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8922,
+      "step": 2215
+    },
+    {
+      "epoch": 0.11300149872954585,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8981,
+      "step": 2216
+    },
+    {
+      "epoch": 0.11305249218565123,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8973,
+      "step": 2217
+    },
+    {
+      "epoch": 0.11310348564175662,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9027,
+      "step": 2218
+    },
+    {
+      "epoch": 0.11315447909786203,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 2219
+    },
+    {
+      "epoch": 0.11320547255396742,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1132564660100728,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8845,
+      "step": 2221
+    },
+    {
+      "epoch": 0.11330745946617819,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 2222
+    },
+    {
+      "epoch": 0.11335845292228358,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8892,
+      "step": 2223
+    },
+    {
+      "epoch": 0.11340944637838897,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 2224
+    },
+    {
+      "epoch": 0.11346043983449436,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.9016,
+      "step": 2225
+    },
+    {
+      "epoch": 0.11351143329059976,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2226
+    },
+    {
+      "epoch": 0.11356242674670515,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8903,
+      "step": 2227
+    },
+    {
+      "epoch": 0.11361342020281054,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8957,
+      "step": 2228
+    },
+    {
+      "epoch": 0.11366441365891593,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 2229
+    },
+    {
+      "epoch": 0.11371540711502132,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1137664005711267,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1138173940272321,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9115,
+      "step": 2232
+    },
+    {
+      "epoch": 0.1138683874833375,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8952,
+      "step": 2233
+    },
+    {
+      "epoch": 0.11391938093944289,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8948,
+      "step": 2234
+    },
+    {
+      "epoch": 0.11397037439554827,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9038,
+      "step": 2235
+    },
+    {
+      "epoch": 0.11402136785165366,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8858,
+      "step": 2236
+    },
+    {
+      "epoch": 0.11407236130775905,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8892,
+      "step": 2237
+    },
+    {
+      "epoch": 0.11412335476386444,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.901,
+      "step": 2238
+    },
+    {
+      "epoch": 0.11417434821996983,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8606,
+      "step": 2239
+    },
+    {
+      "epoch": 0.11422534167607523,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11427633513218062,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.9144,
+      "step": 2241
+    },
+    {
+      "epoch": 0.11432732858828601,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.9054,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1143783220443914,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8801,
+      "step": 2243
+    },
+    {
+      "epoch": 0.11442931550049679,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 2244
+    },
+    {
+      "epoch": 0.11448030895660218,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.9228,
+      "step": 2245
+    },
+    {
+      "epoch": 0.11453130241270756,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8939,
+      "step": 2246
+    },
+    {
+      "epoch": 0.11458229586881297,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 2247
+    },
+    {
+      "epoch": 0.11463328932491836,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8792,
+      "step": 2248
+    },
+    {
+      "epoch": 0.11468428278102374,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 2249
+    },
+    {
+      "epoch": 0.11473527623712913,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11478626969323452,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9073,
+      "step": 2251
+    },
+    {
+      "epoch": 0.11483726314933991,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8763,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1148882566054453,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8942,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1149392500615507,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.908,
+      "step": 2254
+    },
+    {
+      "epoch": 0.11499024351765609,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2255
+    },
+    {
+      "epoch": 0.11504123697376148,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2256
+    },
+    {
+      "epoch": 0.11509223042986687,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8892,
+      "step": 2257
+    },
+    {
+      "epoch": 0.11514322388597226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8921,
+      "step": 2258
+    },
+    {
+      "epoch": 0.11519421734207765,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8992,
+      "step": 2259
+    },
+    {
+      "epoch": 0.11524521079818303,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11529620425428844,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9008,
+      "step": 2261
+    },
+    {
+      "epoch": 0.11534719771039383,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8917,
+      "step": 2262
+    },
+    {
+      "epoch": 0.11539819116649921,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1154491846226046,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9161,
+      "step": 2264
+    },
+    {
+      "epoch": 0.11550017807870999,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8925,
+      "step": 2265
+    },
+    {
+      "epoch": 0.11555117153481538,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8917,
+      "step": 2266
+    },
+    {
+      "epoch": 0.11560216499092077,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8898,
+      "step": 2267
+    },
+    {
+      "epoch": 0.11565315844702617,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 2268
+    },
+    {
+      "epoch": 0.11570415190313156,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2269
+    },
+    {
+      "epoch": 0.11575514535923695,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8854,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11580613881534234,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.897,
+      "step": 2271
+    },
+    {
+      "epoch": 0.11585713227144773,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8997,
+      "step": 2272
+    },
+    {
+      "epoch": 0.11590812572755312,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.9014,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1159591191836585,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 2274
+    },
+    {
+      "epoch": 0.11601011263976391,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9004,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1160611060958693,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9043,
+      "step": 2276
+    },
+    {
+      "epoch": 0.11611209955197468,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9128,
+      "step": 2277
+    },
+    {
+      "epoch": 0.11616309300808007,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 2278
+    },
+    {
+      "epoch": 0.11621408646418546,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.889,
+      "step": 2279
+    },
+    {
+      "epoch": 0.11626507992029085,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11631607337639624,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 2281
+    },
+    {
+      "epoch": 0.11636706683250164,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 2282
+    },
+    {
+      "epoch": 0.11641806028860703,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.9149,
+      "step": 2283
+    },
+    {
+      "epoch": 0.11646905374471242,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8918,
+      "step": 2284
+    },
+    {
+      "epoch": 0.11652004720081781,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1165710406569232,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 2286
+    },
+    {
+      "epoch": 0.11662203411302859,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9185,
+      "step": 2287
+    },
+    {
+      "epoch": 0.11667302756913397,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8878,
+      "step": 2288
+    },
+    {
+      "epoch": 0.11672402102523938,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8964,
+      "step": 2289
+    },
+    {
+      "epoch": 0.11677501448134477,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11682600793745015,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8993,
+      "step": 2291
+    },
+    {
+      "epoch": 0.11687700139355554,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.9043,
+      "step": 2292
+    },
+    {
+      "epoch": 0.11692799484966093,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 2293
+    },
+    {
+      "epoch": 0.11697898830576632,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 2294
+    },
+    {
+      "epoch": 0.11702998176187171,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8992,
+      "step": 2295
+    },
+    {
+      "epoch": 0.11708097521797711,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8773,
+      "step": 2296
+    },
+    {
+      "epoch": 0.1171319686740825,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8951,
+      "step": 2297
+    },
+    {
+      "epoch": 0.11718296213018789,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 2298
+    },
+    {
+      "epoch": 0.11723395558629328,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 2299
+    },
+    {
+      "epoch": 0.11728494904239867,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11733594249850406,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 2301
+    },
+    {
+      "epoch": 0.11738693595460944,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9032,
+      "step": 2302
+    },
+    {
+      "epoch": 0.11743792941071485,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 2303
+    },
+    {
+      "epoch": 0.11748892286682024,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9053,
+      "step": 2304
+    },
+    {
+      "epoch": 0.11753991632292562,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9084,
+      "step": 2305
+    },
+    {
+      "epoch": 0.11759090977903101,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9127,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1176419032351364,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8968,
+      "step": 2307
+    },
+    {
+      "epoch": 0.11769289669124179,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8962,
+      "step": 2308
+    },
+    {
+      "epoch": 0.11774389014734718,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9054,
+      "step": 2309
+    },
+    {
+      "epoch": 0.11779488360345258,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8883,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11784587705955797,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8719,
+      "step": 2311
+    },
+    {
+      "epoch": 0.11789687051566336,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8893,
+      "step": 2312
+    },
+    {
+      "epoch": 0.11794786397176875,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.891,
+      "step": 2313
+    },
+    {
+      "epoch": 0.11799885742787414,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 2314
+    },
+    {
+      "epoch": 0.11804985088397953,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11810084434008492,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 2316
+    },
+    {
+      "epoch": 0.11815183779619032,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9003,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1182028312522957,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8971,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1182538247084011,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.9174,
+      "step": 2319
+    },
+    {
+      "epoch": 0.11830481816450648,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11835581162061187,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9039,
+      "step": 2321
+    },
+    {
+      "epoch": 0.11840680507671726,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8917,
+      "step": 2322
+    },
+    {
+      "epoch": 0.11845779853282265,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 2323
+    },
+    {
+      "epoch": 0.11850879198892805,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9012,
+      "step": 2324
+    },
+    {
+      "epoch": 0.11855978544503344,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11861077890113883,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9109,
+      "step": 2326
+    },
+    {
+      "epoch": 0.11866177235724422,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.9116,
+      "step": 2327
+    },
+    {
+      "epoch": 0.11871276581334961,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8783,
+      "step": 2328
+    },
+    {
+      "epoch": 0.118763759269455,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.9,
+      "step": 2329
+    },
+    {
+      "epoch": 0.11881475272556039,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8893,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11886574618166577,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8942,
+      "step": 2331
+    },
+    {
+      "epoch": 0.11891673963777118,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8946,
+      "step": 2332
+    },
+    {
+      "epoch": 0.11896773309387657,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 2333
+    },
+    {
+      "epoch": 0.11901872654998195,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9077,
+      "step": 2334
+    },
+    {
+      "epoch": 0.11906972000608734,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11912071346219273,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8694,
+      "step": 2336
+    },
+    {
+      "epoch": 0.11917170691829812,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8822,
+      "step": 2337
+    },
+    {
+      "epoch": 0.11922270037440351,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8772,
+      "step": 2338
+    },
+    {
+      "epoch": 0.11927369383050891,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8769,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1193246872866143,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8759,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11937568074271969,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.908,
+      "step": 2341
+    },
+    {
+      "epoch": 0.11942667419882508,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.911,
+      "step": 2342
+    },
+    {
+      "epoch": 0.11947766765493047,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 2343
+    },
+    {
+      "epoch": 0.11952866111103586,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8967,
+      "step": 2344
+    },
+    {
+      "epoch": 0.11957965456714124,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8681,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11963064802324665,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.898,
+      "step": 2346
+    },
+    {
+      "epoch": 0.11968164147935204,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 2347
+    },
+    {
+      "epoch": 0.11973263493545742,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9114,
+      "step": 2348
+    },
+    {
+      "epoch": 0.11978362839156281,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.895,
+      "step": 2349
+    },
+    {
+      "epoch": 0.1198346218476682,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8997,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11988561530377359,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 2351
+    },
+    {
+      "epoch": 0.11993660875987898,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 2352
+    },
+    {
+      "epoch": 0.11998760221598438,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 2353
+    },
+    {
+      "epoch": 0.12003859567208977,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8848,
+      "step": 2354
+    },
+    {
+      "epoch": 0.12008958912819516,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 2355
+    },
+    {
+      "epoch": 0.12014058258430055,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12019157604040594,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9222,
+      "step": 2357
+    },
+    {
+      "epoch": 0.12024256949651133,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2358
+    },
+    {
+      "epoch": 0.12029356295261671,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 2359
+    },
+    {
+      "epoch": 0.12034455640872212,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1203955498648275,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1204465433209329,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.9094,
+      "step": 2362
+    },
+    {
+      "epoch": 0.12049753677703828,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9011,
+      "step": 2363
+    },
+    {
+      "epoch": 0.12054853023314367,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9014,
+      "step": 2364
+    },
+    {
+      "epoch": 0.12059952368924906,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 2365
+    },
+    {
+      "epoch": 0.12065051714535445,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9244,
+      "step": 2366
+    },
+    {
+      "epoch": 0.12070151060145985,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.905,
+      "step": 2367
+    },
+    {
+      "epoch": 0.12075250405756524,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 2368
+    },
+    {
+      "epoch": 0.12080349751367063,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 2369
+    },
+    {
+      "epoch": 0.12085449096977602,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12090548442588141,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 2371
+    },
+    {
+      "epoch": 0.1209564778819868,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9046,
+      "step": 2372
+    },
+    {
+      "epoch": 0.12100747133809218,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8838,
+      "step": 2373
+    },
+    {
+      "epoch": 0.12105846479419759,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 2374
+    },
+    {
+      "epoch": 0.12110945825030298,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8825,
+      "step": 2375
+    },
+    {
+      "epoch": 0.12116045170640836,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 2376
+    },
+    {
+      "epoch": 0.12121144516251375,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9033,
+      "step": 2377
+    },
+    {
+      "epoch": 0.12126243861861914,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8867,
+      "step": 2378
+    },
+    {
+      "epoch": 0.12131343207472453,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 2379
+    },
+    {
+      "epoch": 0.12136442553082992,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12141541898693532,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9165,
+      "step": 2381
+    },
+    {
+      "epoch": 0.12146641244304071,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 2382
+    },
+    {
+      "epoch": 0.1215174058991461,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.891,
+      "step": 2383
+    },
+    {
+      "epoch": 0.12156839935525149,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9005,
+      "step": 2384
+    },
+    {
+      "epoch": 0.12161939281135688,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 2385
+    },
+    {
+      "epoch": 0.12167038626746227,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 2386
+    },
+    {
+      "epoch": 0.12172137972356765,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9002,
+      "step": 2387
+    },
+    {
+      "epoch": 0.12177237317967306,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8815,
+      "step": 2388
+    },
+    {
+      "epoch": 0.12182336663577845,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.9016,
+      "step": 2389
+    },
+    {
+      "epoch": 0.12187436009188383,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12192535354798922,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8813,
+      "step": 2391
+    },
+    {
+      "epoch": 0.12197634700409461,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1220273404602,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 2393
+    },
+    {
+      "epoch": 0.12207833391630539,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 2394
+    },
+    {
+      "epoch": 0.12212932737241079,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8881,
+      "step": 2395
+    },
+    {
+      "epoch": 0.12218032082851618,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 2396
+    },
+    {
+      "epoch": 0.12223131428462157,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8866,
+      "step": 2397
+    },
+    {
+      "epoch": 0.12228230774072696,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 2398
+    },
+    {
+      "epoch": 0.12233330119683235,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 2399
+    },
+    {
+      "epoch": 0.12238429465293774,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.87,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12243528810904312,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2401
+    },
+    {
+      "epoch": 0.12248628156514853,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 2402
+    },
+    {
+      "epoch": 0.12253727502125392,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1225882684773593,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8841,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1226392619334647,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9032,
+      "step": 2405
+    },
+    {
+      "epoch": 0.12269025538957008,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.9054,
+      "step": 2406
+    },
+    {
+      "epoch": 0.12274124884567547,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8882,
+      "step": 2407
+    },
+    {
+      "epoch": 0.12279224230178086,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 2408
+    },
+    {
+      "epoch": 0.12284323575788626,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8906,
+      "step": 2409
+    },
+    {
+      "epoch": 0.12289422921399165,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8921,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12294522267009704,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 2411
+    },
+    {
+      "epoch": 0.12299621612620243,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 2412
+    },
+    {
+      "epoch": 0.12304720958230782,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 2413
+    },
+    {
+      "epoch": 0.1230982030384132,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1231491964945186,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.913,
+      "step": 2415
+    },
+    {
+      "epoch": 0.123200189950624,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.9045,
+      "step": 2416
+    },
+    {
+      "epoch": 0.12325118340672939,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8942,
+      "step": 2417
+    },
+    {
+      "epoch": 0.12330217686283478,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.894,
+      "step": 2418
+    },
+    {
+      "epoch": 0.12335317031894016,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9026,
+      "step": 2419
+    },
+    {
+      "epoch": 0.12340416377504555,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12345515723115094,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9021,
+      "step": 2421
+    },
+    {
+      "epoch": 0.12350615068725633,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 2422
+    },
+    {
+      "epoch": 0.12355714414336173,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8884,
+      "step": 2423
+    },
+    {
+      "epoch": 0.12360813759946712,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 2424
+    },
+    {
+      "epoch": 0.12365913105557251,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.9019,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1237101245116779,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8774,
+      "step": 2426
+    },
+    {
+      "epoch": 0.12376111796778329,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 2427
+    },
+    {
+      "epoch": 0.12381211142388868,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.9132,
+      "step": 2428
+    },
+    {
+      "epoch": 0.12386310487999407,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8956,
+      "step": 2429
+    },
+    {
+      "epoch": 0.12391409833609947,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8959,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12396509179220486,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9207,
+      "step": 2431
+    },
+    {
+      "epoch": 0.12401608524831025,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8798,
+      "step": 2432
+    },
+    {
+      "epoch": 0.12406707870441563,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8986,
+      "step": 2433
+    },
+    {
+      "epoch": 0.12411807216052102,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8863,
+      "step": 2434
+    },
+    {
+      "epoch": 0.12416906561662641,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1242200590727318,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 2436
+    },
+    {
+      "epoch": 0.1242710525288372,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 2437
+    },
+    {
+      "epoch": 0.12432204598494259,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 2438
+    },
+    {
+      "epoch": 0.12437303944104798,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 2439
+    },
+    {
+      "epoch": 0.12442403289715337,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9036,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12447502635325876,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8881,
+      "step": 2441
+    },
+    {
+      "epoch": 0.12452601980936415,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8956,
+      "step": 2442
+    },
+    {
+      "epoch": 0.12457701326546954,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8906,
+      "step": 2443
+    },
+    {
+      "epoch": 0.12462800672157494,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8743,
+      "step": 2444
+    },
+    {
+      "epoch": 0.12467900017768033,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8824,
+      "step": 2445
+    },
+    {
+      "epoch": 0.12472999363378572,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 2446
+    },
+    {
+      "epoch": 0.1247809870898911,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8982,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1248319805459965,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 2448
+    },
+    {
+      "epoch": 0.12488297400210188,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.877,
+      "step": 2449
+    },
+    {
+      "epoch": 0.12493396745820727,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8974,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12498496091431267,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 2451
+    },
+    {
+      "epoch": 0.12503595437041806,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8797,
+      "step": 2452
+    },
+    {
+      "epoch": 0.12508694782652344,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8789,
+      "step": 2453
+    },
+    {
+      "epoch": 0.12513794128262884,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 2454
+    },
+    {
+      "epoch": 0.12518893473873424,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 2455
+    },
+    {
+      "epoch": 0.12523992819483962,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 2456
+    },
+    {
+      "epoch": 0.12529092165094502,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9129,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1253419151070504,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8877,
+      "step": 2458
+    },
+    {
+      "epoch": 0.1253929085631558,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 2459
+    },
+    {
+      "epoch": 0.12544390201926117,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8782,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12549489547536657,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8783,
+      "step": 2461
+    },
+    {
+      "epoch": 0.12554588893147198,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.9116,
+      "step": 2462
+    },
+    {
+      "epoch": 0.12559688238757735,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2463
+    },
+    {
+      "epoch": 0.12564787584368275,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8969,
+      "step": 2464
+    },
+    {
+      "epoch": 0.12569886929978813,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8853,
+      "step": 2465
+    },
+    {
+      "epoch": 0.12574986275589353,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 2466
+    },
+    {
+      "epoch": 0.1258008562119989,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.896,
+      "step": 2467
+    },
+    {
+      "epoch": 0.1258518496681043,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.904,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1259028431242097,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9018,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1259538365803151,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1260048300364205,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 2471
+    },
+    {
+      "epoch": 0.12605582349252586,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 2472
+    },
+    {
+      "epoch": 0.12610681694863127,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9128,
+      "step": 2473
+    },
+    {
+      "epoch": 0.12615781040473664,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8994,
+      "step": 2474
+    },
+    {
+      "epoch": 0.12620880386084204,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 2475
+    },
+    {
+      "epoch": 0.12625979731694745,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8836,
+      "step": 2476
+    },
+    {
+      "epoch": 0.12631079077305282,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 2477
+    },
+    {
+      "epoch": 0.12636178422915822,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8963,
+      "step": 2478
+    },
+    {
+      "epoch": 0.1264127776852636,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8832,
+      "step": 2479
+    },
+    {
+      "epoch": 0.126463771141369,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8817,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12651476459747438,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8907,
+      "step": 2481
+    },
+    {
+      "epoch": 0.12656575805357978,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 2482
+    },
+    {
+      "epoch": 0.12661675150968518,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.904,
+      "step": 2483
+    },
+    {
+      "epoch": 0.12666774496579056,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 2484
+    },
+    {
+      "epoch": 0.12671873842189596,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 2485
+    },
+    {
+      "epoch": 0.12676973187800133,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8725,
+      "step": 2486
+    },
+    {
+      "epoch": 0.12682072533410674,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1268717187902121,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.9013,
+      "step": 2488
+    },
+    {
+      "epoch": 0.12692271224631752,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2489
+    },
+    {
+      "epoch": 0.12697370570242292,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8915,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1270246991585283,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 2491
+    },
+    {
+      "epoch": 0.1270756926146337,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 2492
+    },
+    {
+      "epoch": 0.12712668607073907,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 2493
+    },
+    {
+      "epoch": 0.12717767952684447,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 2494
+    },
+    {
+      "epoch": 0.12722867298294985,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8929,
+      "step": 2495
+    },
+    {
+      "epoch": 0.12727966643905525,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9116,
+      "step": 2496
+    },
+    {
+      "epoch": 0.12733065989516065,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8832,
+      "step": 2497
+    },
+    {
+      "epoch": 0.12738165335126603,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 2498
+    },
+    {
+      "epoch": 0.12743264680737143,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 2499
+    },
+    {
+      "epoch": 0.1274836402634768,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.896,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1275346337195822,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8759,
+      "step": 2501
+    },
+    {
+      "epoch": 0.12758562717568758,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8825,
+      "step": 2502
+    },
+    {
+      "epoch": 0.12763662063179299,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.9027,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1276876140878984,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.9006,
+      "step": 2504
+    },
+    {
+      "epoch": 0.12773860754400376,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8913,
+      "step": 2505
+    },
+    {
+      "epoch": 0.12778960100010917,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 2506
+    },
+    {
+      "epoch": 0.12784059445621454,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8886,
+      "step": 2507
+    },
+    {
+      "epoch": 0.12789158791231994,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 2508
+    },
+    {
+      "epoch": 0.12794258136842532,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 2509
+    },
+    {
+      "epoch": 0.12799357482453072,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12804456828063612,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8833,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1280955617367415,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8942,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1281465551928469,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 2513
+    },
+    {
+      "epoch": 0.12819754864895228,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 2514
+    },
+    {
+      "epoch": 0.12824854210505768,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8884,
+      "step": 2515
+    },
+    {
+      "epoch": 0.12829953556116305,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8774,
+      "step": 2516
+    },
+    {
+      "epoch": 0.12835052901726846,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2517
+    },
+    {
+      "epoch": 0.12840152247337386,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8794,
+      "step": 2518
+    },
+    {
+      "epoch": 0.12845251592947923,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 2519
+    },
+    {
+      "epoch": 0.12850350938558464,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8926,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12855450284169,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8948,
+      "step": 2521
+    },
+    {
+      "epoch": 0.1286054962977954,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9074,
+      "step": 2522
+    },
+    {
+      "epoch": 0.1286564897539008,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1287074832100062,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.883,
+      "step": 2524
+    },
+    {
+      "epoch": 0.1287584766661116,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12880947012221697,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 2526
+    },
+    {
+      "epoch": 0.12886046357832237,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 2527
+    },
+    {
+      "epoch": 0.12891145703442775,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8792,
+      "step": 2528
+    },
+    {
+      "epoch": 0.12896245049053315,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9272,
+      "step": 2529
+    },
+    {
+      "epoch": 0.12901344394663852,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.904,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12906443740274393,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8933,
+      "step": 2531
+    },
+    {
+      "epoch": 0.12911543085884933,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.9038,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1291664243149547,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8607,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1292174177710601,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.9193,
+      "step": 2534
+    },
+    {
+      "epoch": 0.12926841122716548,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8825,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12931940468327088,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8928,
+      "step": 2536
+    },
+    {
+      "epoch": 0.12937039813937626,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8783,
+      "step": 2537
+    },
+    {
+      "epoch": 0.12942139159548166,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 2538
+    },
+    {
+      "epoch": 0.12947238505158706,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8737,
+      "step": 2539
+    },
+    {
+      "epoch": 0.12952337850769244,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12957437196379784,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.895,
+      "step": 2541
+    },
+    {
+      "epoch": 0.12962536541990322,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2542
+    },
+    {
+      "epoch": 0.12967635887600862,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 2543
+    },
+    {
+      "epoch": 0.129727352332114,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.9031,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1297783457882194,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1298293392443248,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8994,
+      "step": 2546
+    },
+    {
+      "epoch": 0.12988033270043017,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.864,
+      "step": 2547
+    },
+    {
+      "epoch": 0.12993132615653558,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.874,
+      "step": 2548
+    },
+    {
+      "epoch": 0.12998231961264095,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9063,
+      "step": 2549
+    },
+    {
+      "epoch": 0.13003331306874635,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8785,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13008430652485173,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 2551
+    },
+    {
+      "epoch": 0.13013529998095713,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13018629343706253,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 2553
+    },
+    {
+      "epoch": 0.1302372868931679,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1302882803492733,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 2555
+    },
+    {
+      "epoch": 0.13033927380537869,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8569,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1303902672614841,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8971,
+      "step": 2557
+    },
+    {
+      "epoch": 0.13044126071758946,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.9078,
+      "step": 2558
+    },
+    {
+      "epoch": 0.13049225417369487,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8882,
+      "step": 2559
+    },
+    {
+      "epoch": 0.13054324762980027,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13059424108590564,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 2561
+    },
+    {
+      "epoch": 0.13064523454201105,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8917,
+      "step": 2562
+    },
+    {
+      "epoch": 0.13069622799811642,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.9022,
+      "step": 2563
+    },
+    {
+      "epoch": 0.13074722145422182,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8939,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1307982149103272,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1308492083664326,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8897,
+      "step": 2566
+    },
+    {
+      "epoch": 0.130900201822538,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 2567
+    },
+    {
+      "epoch": 0.13095119527864338,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8825,
+      "step": 2568
+    },
+    {
+      "epoch": 0.13100218873474878,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 2569
+    },
+    {
+      "epoch": 0.13105318219085416,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.9084,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13110417564695956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.9042,
+      "step": 2571
+    },
+    {
+      "epoch": 0.13115516910306493,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8875,
+      "step": 2572
+    },
+    {
+      "epoch": 0.13120616255917034,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8761,
+      "step": 2573
+    },
+    {
+      "epoch": 0.13125715601527574,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8927,
+      "step": 2574
+    },
+    {
+      "epoch": 0.1313081494713811,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2575
+    },
+    {
+      "epoch": 0.13135914292748652,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8983,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1314101363835919,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8936,
+      "step": 2577
+    },
+    {
+      "epoch": 0.1314611298396973,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 2578
+    },
+    {
+      "epoch": 0.13151212329580267,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8956,
+      "step": 2579
+    },
+    {
+      "epoch": 0.13156311675190807,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13161411020801347,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9026,
+      "step": 2581
+    },
+    {
+      "epoch": 0.13166510366411885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 2582
+    },
+    {
+      "epoch": 0.13171609712022425,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8811,
+      "step": 2583
+    },
+    {
+      "epoch": 0.13176709057632963,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8899,
+      "step": 2584
+    },
+    {
+      "epoch": 0.13181808403243503,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8557,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1318690774885404,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 2586
+    },
+    {
+      "epoch": 0.1319200709446458,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8869,
+      "step": 2587
+    },
+    {
+      "epoch": 0.1319710644007512,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9106,
+      "step": 2588
+    },
+    {
+      "epoch": 0.13202205785685658,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8816,
+      "step": 2589
+    },
+    {
+      "epoch": 0.132073051312962,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13212404476906736,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 2591
+    },
+    {
+      "epoch": 0.13217503822517276,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 2592
+    },
+    {
+      "epoch": 0.13222603168127814,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 2593
+    },
+    {
+      "epoch": 0.13227702513738354,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 2594
+    },
+    {
+      "epoch": 0.13232801859348894,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.9039,
+      "step": 2595
+    },
+    {
+      "epoch": 0.13237901204959432,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 2596
+    },
+    {
+      "epoch": 0.13243000550569972,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1324809989618051,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8984,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1325319924179105,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 2599
+    },
+    {
+      "epoch": 0.13258298587401587,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13263397933012128,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.9157,
+      "step": 2601
+    },
+    {
+      "epoch": 0.13268497278622668,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 2602
+    },
+    {
+      "epoch": 0.13273596624233205,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 2603
+    },
+    {
+      "epoch": 0.13278695969843746,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 2604
+    },
+    {
+      "epoch": 0.13283795315454283,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8997,
+      "step": 2605
+    },
+    {
+      "epoch": 0.13288894661064823,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 2606
+    },
+    {
+      "epoch": 0.1329399400667536,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 2607
+    },
+    {
+      "epoch": 0.132990933522859,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 2608
+    },
+    {
+      "epoch": 0.13304192697896441,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.883,
+      "step": 2609
+    },
+    {
+      "epoch": 0.1330929204350698,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8869,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1331439138911752,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.893,
+      "step": 2611
+    },
+    {
+      "epoch": 0.13319490734728057,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8839,
+      "step": 2612
+    },
+    {
+      "epoch": 0.13324590080338597,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 2613
+    },
+    {
+      "epoch": 0.13329689425949134,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8871,
+      "step": 2614
+    },
+    {
+      "epoch": 0.13334788771559675,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 2615
+    },
+    {
+      "epoch": 0.13339888117170215,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8932,
+      "step": 2616
+    },
+    {
+      "epoch": 0.13344987462780752,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 2617
+    },
+    {
+      "epoch": 0.13350086808391293,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.9012,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1335518615400183,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 2619
+    },
+    {
+      "epoch": 0.1336028549961237,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13365384845222908,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8954,
+      "step": 2621
+    },
+    {
+      "epoch": 0.13370484190833448,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 2622
+    },
+    {
+      "epoch": 0.13375583536443988,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8768,
+      "step": 2623
+    },
+    {
+      "epoch": 0.13380682882054526,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 2624
+    },
+    {
+      "epoch": 0.13385782227665066,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8772,
+      "step": 2625
+    },
+    {
+      "epoch": 0.13390881573275604,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 2626
+    },
+    {
+      "epoch": 0.13395980918886144,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8871,
+      "step": 2627
+    },
+    {
+      "epoch": 0.13401080264496681,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9196,
+      "step": 2628
+    },
+    {
+      "epoch": 0.13406179610107222,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 2629
+    },
+    {
+      "epoch": 0.13411278955717762,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 2630
+    },
+    {
+      "epoch": 0.134163783013283,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9039,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1342147764693884,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 2632
+    },
+    {
+      "epoch": 0.13426576992549377,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 2633
+    },
+    {
+      "epoch": 0.13431676338159917,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8624,
+      "step": 2634
+    },
+    {
+      "epoch": 0.13436775683770455,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 2635
+    },
+    {
+      "epoch": 0.13441875029380995,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 2636
+    },
+    {
+      "epoch": 0.13446974374991535,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 2637
+    },
+    {
+      "epoch": 0.13452073720602073,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8808,
+      "step": 2638
+    },
+    {
+      "epoch": 0.13457173066212613,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8901,
+      "step": 2639
+    },
+    {
+      "epoch": 0.1346227241182315,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1346737175743369,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 2641
+    },
+    {
+      "epoch": 0.13472471103044228,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 2642
+    },
+    {
+      "epoch": 0.1347757044865477,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 2643
+    },
+    {
+      "epoch": 0.1348266979426531,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 2644
+    },
+    {
+      "epoch": 0.13487769139875846,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8846,
+      "step": 2645
+    },
+    {
+      "epoch": 0.13492868485486387,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 2646
+    },
+    {
+      "epoch": 0.13497967831096924,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 2647
+    },
+    {
+      "epoch": 0.13503067176707464,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8799,
+      "step": 2648
+    },
+    {
+      "epoch": 0.13508166522318002,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 2649
+    },
+    {
+      "epoch": 0.13513265867928542,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8859,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13518365213539082,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 2651
+    },
+    {
+      "epoch": 0.1352346455914962,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 2652
+    },
+    {
+      "epoch": 0.1352856390476016,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 2653
+    },
+    {
+      "epoch": 0.13533663250370698,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 2654
+    },
+    {
+      "epoch": 0.13538762595981238,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8899,
+      "step": 2655
+    },
+    {
+      "epoch": 0.13543861941591775,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 2656
+    },
+    {
+      "epoch": 0.13548961287202316,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8988,
+      "step": 2657
+    },
+    {
+      "epoch": 0.13554060632812856,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 2658
+    },
+    {
+      "epoch": 0.13559159978423393,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8857,
+      "step": 2659
+    },
+    {
+      "epoch": 0.13564259324033934,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1356935866964447,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8841,
+      "step": 2661
+    },
+    {
+      "epoch": 0.13574458015255012,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1357955736086555,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 2663
+    },
+    {
+      "epoch": 0.1358465670647609,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 2664
+    },
+    {
+      "epoch": 0.1358975605208663,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8873,
+      "step": 2665
+    },
+    {
+      "epoch": 0.13594855397697167,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8913,
+      "step": 2666
+    },
+    {
+      "epoch": 0.13599954743307707,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8671,
+      "step": 2667
+    },
+    {
+      "epoch": 0.13605054088918245,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 2668
+    },
+    {
+      "epoch": 0.13610153434528785,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.9027,
+      "step": 2669
+    },
+    {
+      "epoch": 0.13615252780139322,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 2670
+    },
+    {
+      "epoch": 0.13620352125749863,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8829,
+      "step": 2671
+    },
+    {
+      "epoch": 0.13625451471360403,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 2672
+    },
+    {
+      "epoch": 0.1363055081697094,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8964,
+      "step": 2673
+    },
+    {
+      "epoch": 0.1363565016258148,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8716,
+      "step": 2674
+    },
+    {
+      "epoch": 0.13640749508192018,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8813,
+      "step": 2675
+    },
+    {
+      "epoch": 0.13645848853802559,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8898,
+      "step": 2676
+    },
+    {
+      "epoch": 0.13650948199413096,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.881,
+      "step": 2677
+    },
+    {
+      "epoch": 0.13656047545023636,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 2678
+    },
+    {
+      "epoch": 0.13661146890634177,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 2679
+    },
+    {
+      "epoch": 0.13666246236244714,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13671345581855254,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.885,
+      "step": 2681
+    },
+    {
+      "epoch": 0.13676444927465792,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9188,
+      "step": 2682
+    },
+    {
+      "epoch": 0.13681544273076332,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1368664361868687,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 2684
+    },
+    {
+      "epoch": 0.1369174296429741,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.87,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1369684230990795,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 2686
+    },
+    {
+      "epoch": 0.13701941655518488,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 2687
+    },
+    {
+      "epoch": 0.13707041001129028,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8843,
+      "step": 2688
+    },
+    {
+      "epoch": 0.13712140346739565,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 2689
+    },
+    {
+      "epoch": 0.13717239692350106,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 2690
+    },
+    {
+      "epoch": 0.13722339037960643,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 2691
+    },
+    {
+      "epoch": 0.13727438383571183,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8986,
+      "step": 2692
+    },
+    {
+      "epoch": 0.13732537729181724,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8907,
+      "step": 2693
+    },
+    {
+      "epoch": 0.1373763707479226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.887,
+      "step": 2694
+    },
+    {
+      "epoch": 0.137427364204028,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1374783576601334,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8814,
+      "step": 2696
+    },
+    {
+      "epoch": 0.1375293511162388,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 2697
+    },
+    {
+      "epoch": 0.13758034457234417,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 2698
+    },
+    {
+      "epoch": 0.13763133802844957,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 2699
+    },
+    {
+      "epoch": 0.13768233148455497,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13773332494066035,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 2701
+    },
+    {
+      "epoch": 0.13778431839676575,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8843,
+      "step": 2702
+    },
+    {
+      "epoch": 0.13783531185287112,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2703
+    },
+    {
+      "epoch": 0.13788630530897653,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8882,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1379372987650819,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.881,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1379882922211873,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.894,
+      "step": 2706
+    },
+    {
+      "epoch": 0.1380392856772927,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8908,
+      "step": 2707
+    },
+    {
+      "epoch": 0.13809027913339808,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8867,
+      "step": 2708
+    },
+    {
+      "epoch": 0.13814127258950348,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8912,
+      "step": 2709
+    },
+    {
+      "epoch": 0.13819226604560886,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13824325950171426,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 2711
+    },
+    {
+      "epoch": 0.13829425295781964,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8907,
+      "step": 2712
+    },
+    {
+      "epoch": 0.13834524641392504,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 2713
+    },
+    {
+      "epoch": 0.13839623987003044,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 2714
+    },
+    {
+      "epoch": 0.13844723332613582,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 2715
+    },
+    {
+      "epoch": 0.13849822678224122,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1385492202383466,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.886,
+      "step": 2717
+    },
+    {
+      "epoch": 0.138600213694452,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 2718
+    },
+    {
+      "epoch": 0.13865120715055737,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8673,
+      "step": 2719
+    },
+    {
+      "epoch": 0.13870220060666277,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8915,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13875319406276818,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8938,
+      "step": 2721
+    },
+    {
+      "epoch": 0.13880418751887355,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8757,
+      "step": 2722
+    },
+    {
+      "epoch": 0.13885518097497895,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8905,
+      "step": 2723
+    },
+    {
+      "epoch": 0.13890617443108433,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8869,
+      "step": 2724
+    },
+    {
+      "epoch": 0.13895716788718973,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8875,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1390081613432951,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.88,
+      "step": 2726
+    },
+    {
+      "epoch": 0.1390591547994005,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8977,
+      "step": 2727
+    },
+    {
+      "epoch": 0.1391101482555059,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 2728
+    },
+    {
+      "epoch": 0.13916114171161129,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.9045,
+      "step": 2729
+    },
+    {
+      "epoch": 0.1392121351677167,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8975,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13926312862382206,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8739,
+      "step": 2731
+    },
+    {
+      "epoch": 0.13931412207992747,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8762,
+      "step": 2732
+    },
+    {
+      "epoch": 0.13936511553603284,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2733
+    },
+    {
+      "epoch": 0.13941610899213824,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8936,
+      "step": 2734
+    },
+    {
+      "epoch": 0.13946710244824365,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8868,
+      "step": 2735
+    },
+    {
+      "epoch": 0.13951809590434902,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 2736
+    },
+    {
+      "epoch": 0.13956908936045442,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.893,
+      "step": 2737
+    },
+    {
+      "epoch": 0.1396200828165598,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 2738
+    },
+    {
+      "epoch": 0.1396710762726652,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8587,
+      "step": 2739
+    },
+    {
+      "epoch": 0.13972206972877058,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13977306318487598,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 2741
+    },
+    {
+      "epoch": 0.13982405664098138,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 2742
+    },
+    {
+      "epoch": 0.13987505009708676,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 2743
+    },
+    {
+      "epoch": 0.13992604355319216,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 2744
+    },
+    {
+      "epoch": 0.13997703700929753,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8841,
+      "step": 2745
+    },
+    {
+      "epoch": 0.14002803046540294,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 2746
+    },
+    {
+      "epoch": 0.1400790239215083,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.901,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1401300173776137,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.88,
+      "step": 2748
+    },
+    {
+      "epoch": 0.14018101083371912,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 2749
+    },
+    {
+      "epoch": 0.1402320042898245,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8974,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1402829977459299,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2751
+    },
+    {
+      "epoch": 0.14033399120203527,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 2752
+    },
+    {
+      "epoch": 0.14038498465814067,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 2753
+    },
+    {
+      "epoch": 0.14043597811424605,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 2754
+    },
+    {
+      "epoch": 0.14048697157035145,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.9164,
+      "step": 2755
+    },
+    {
+      "epoch": 0.14053796502645685,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 2756
+    },
+    {
+      "epoch": 0.14058895848256223,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 2757
+    },
+    {
+      "epoch": 0.14063995193866763,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8966,
+      "step": 2758
+    },
+    {
+      "epoch": 0.140690945394773,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 2759
+    },
+    {
+      "epoch": 0.1407419388508784,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14079293230698378,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8938,
+      "step": 2761
+    },
+    {
+      "epoch": 0.14084392576308918,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.88,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1408949192191946,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8606,
+      "step": 2763
+    },
+    {
+      "epoch": 0.14094591267529996,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 2764
+    },
+    {
+      "epoch": 0.14099690613140536,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8895,
+      "step": 2765
+    },
+    {
+      "epoch": 0.14104789958751074,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 2766
+    },
+    {
+      "epoch": 0.14109889304361614,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8936,
+      "step": 2767
+    },
+    {
+      "epoch": 0.14114988649972152,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8713,
+      "step": 2768
+    },
+    {
+      "epoch": 0.14120087995582692,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 2769
+    },
+    {
+      "epoch": 0.14125187341193232,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.874,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1413028668680377,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2771
+    },
+    {
+      "epoch": 0.1413538603241431,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 2772
+    },
+    {
+      "epoch": 0.14140485378024847,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8947,
+      "step": 2773
+    },
+    {
+      "epoch": 0.14145584723635388,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 2774
+    },
+    {
+      "epoch": 0.14150684069245925,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 2775
+    },
+    {
+      "epoch": 0.14155783414856465,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 2776
+    },
+    {
+      "epoch": 0.14160882760467006,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8792,
+      "step": 2777
+    },
+    {
+      "epoch": 0.14165982106077543,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 2778
+    },
+    {
+      "epoch": 0.14171081451688083,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 2779
+    },
+    {
+      "epoch": 0.1417618079729862,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1418128014290916,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 2781
+    },
+    {
+      "epoch": 0.141863794885197,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 2782
+    },
+    {
+      "epoch": 0.1419147883413024,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1419657817974078,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 2784
+    },
+    {
+      "epoch": 0.14201677525351317,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8848,
+      "step": 2785
+    },
+    {
+      "epoch": 0.14206776870961857,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 2786
+    },
+    {
+      "epoch": 0.14211876216572394,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8727,
+      "step": 2787
+    },
+    {
+      "epoch": 0.14216975562182935,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8822,
+      "step": 2788
+    },
+    {
+      "epoch": 0.14222074907793472,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8857,
+      "step": 2789
+    },
+    {
+      "epoch": 0.14227174253404012,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.895,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14232273599014553,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 2791
+    },
+    {
+      "epoch": 0.1423737294462509,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8933,
+      "step": 2792
+    },
+    {
+      "epoch": 0.1424247229023563,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 2793
+    },
+    {
+      "epoch": 0.14247571635846168,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 2794
+    },
+    {
+      "epoch": 0.14252670981456708,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8989,
+      "step": 2795
+    },
+    {
+      "epoch": 0.14257770327067246,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 2796
+    },
+    {
+      "epoch": 0.14262869672677786,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 2797
+    },
+    {
+      "epoch": 0.14267969018288326,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 2798
+    },
+    {
+      "epoch": 0.14273068363898864,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 2799
+    },
+    {
+      "epoch": 0.14278167709509404,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14283267055119941,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 2801
+    },
+    {
+      "epoch": 0.14288366400730482,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1429346574634102,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 2803
+    },
+    {
+      "epoch": 0.1429856509195156,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 2804
+    },
+    {
+      "epoch": 0.143036644375621,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8948,
+      "step": 2805
+    },
+    {
+      "epoch": 0.14308763783172637,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.9148,
+      "step": 2806
+    },
+    {
+      "epoch": 0.14313863128783177,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 2807
+    },
+    {
+      "epoch": 0.14318962474393715,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.899,
+      "step": 2808
+    },
+    {
+      "epoch": 0.14324061820004255,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 2809
+    },
+    {
+      "epoch": 0.14329161165614793,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14334260511225333,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8846,
+      "step": 2811
+    },
+    {
+      "epoch": 0.14339359856835873,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.936,
+      "step": 2812
+    },
+    {
+      "epoch": 0.1434445920244641,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2813
+    },
+    {
+      "epoch": 0.1434955854805695,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 2814
+    },
+    {
+      "epoch": 0.14354657893667488,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.846,
+      "step": 2815
+    },
+    {
+      "epoch": 0.1435975723927803,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 2816
+    },
+    {
+      "epoch": 0.14364856584888566,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9023,
+      "step": 2817
+    },
+    {
+      "epoch": 0.14369955930499106,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8811,
+      "step": 2818
+    },
+    {
+      "epoch": 0.14375055276109647,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 2819
+    },
+    {
+      "epoch": 0.14380154621720184,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14385253967330724,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 2821
+    },
+    {
+      "epoch": 0.14390353312941262,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8681,
+      "step": 2822
+    },
+    {
+      "epoch": 0.14395452658551802,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8926,
+      "step": 2823
+    },
+    {
+      "epoch": 0.1440055200416234,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1440565134977288,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 2825
+    },
+    {
+      "epoch": 0.1441075069538342,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8921,
+      "step": 2826
+    },
+    {
+      "epoch": 0.14415850040993958,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8782,
+      "step": 2827
+    },
+    {
+      "epoch": 0.14420949386604498,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8863,
+      "step": 2828
+    },
+    {
+      "epoch": 0.14426048732215035,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 2829
+    },
+    {
+      "epoch": 0.14431148077825576,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8912,
+      "step": 2830
+    },
+    {
+      "epoch": 0.14436247423436113,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 2831
+    },
+    {
+      "epoch": 0.14441346769046653,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8701,
+      "step": 2832
+    },
+    {
+      "epoch": 0.14446446114657194,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2833
+    },
+    {
+      "epoch": 0.1445154546026773,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2834
+    },
+    {
+      "epoch": 0.14456644805878271,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 2835
+    },
+    {
+      "epoch": 0.1446174415148881,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8645,
+      "step": 2836
+    },
+    {
+      "epoch": 0.1446684349709935,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 2837
+    },
+    {
+      "epoch": 0.14471942842709887,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 2838
+    },
+    {
+      "epoch": 0.14477042188320427,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8679,
+      "step": 2839
+    },
+    {
+      "epoch": 0.14482141533930967,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8853,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14487240879541505,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8929,
+      "step": 2841
+    },
+    {
+      "epoch": 0.14492340225152045,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.9121,
+      "step": 2842
+    },
+    {
+      "epoch": 0.14497439570762582,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8915,
+      "step": 2843
+    },
+    {
+      "epoch": 0.14502538916373123,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 2844
+    },
+    {
+      "epoch": 0.1450763826198366,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8719,
+      "step": 2845
+    },
+    {
+      "epoch": 0.145127376075942,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 2846
+    },
+    {
+      "epoch": 0.1451783695320474,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8871,
+      "step": 2847
+    },
+    {
+      "epoch": 0.14522936298815278,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 2848
+    },
+    {
+      "epoch": 0.14528035644425819,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 2849
+    },
+    {
+      "epoch": 0.14533134990036356,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9003,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14538234335646896,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2851
+    },
+    {
+      "epoch": 0.14543333681257434,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8977,
+      "step": 2852
+    },
+    {
+      "epoch": 0.14548433026867974,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8908,
+      "step": 2853
+    },
+    {
+      "epoch": 0.14553532372478514,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 2854
+    },
+    {
+      "epoch": 0.14558631718089052,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2855
+    },
+    {
+      "epoch": 0.14563731063699592,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 2856
+    },
+    {
+      "epoch": 0.1456883040931013,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8799,
+      "step": 2857
+    },
+    {
+      "epoch": 0.1457392975492067,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8838,
+      "step": 2858
+    },
+    {
+      "epoch": 0.14579029100531207,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 2859
+    },
+    {
+      "epoch": 0.14584128446141748,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14589227791752288,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8953,
+      "step": 2861
+    },
+    {
+      "epoch": 0.14594327137362825,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8797,
+      "step": 2862
+    },
+    {
+      "epoch": 0.14599426482973366,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 2863
+    },
+    {
+      "epoch": 0.14604525828583903,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 2864
+    },
+    {
+      "epoch": 0.14609625174194443,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8926,
+      "step": 2865
+    },
+    {
+      "epoch": 0.1461472451980498,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 2866
+    },
+    {
+      "epoch": 0.1461982386541552,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8758,
+      "step": 2867
+    },
+    {
+      "epoch": 0.1462492321102606,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 2868
+    },
+    {
+      "epoch": 0.146300225566366,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1463512190224714,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8914,
+      "step": 2870
+    },
+    {
+      "epoch": 0.14640221247857677,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 2871
+    },
+    {
+      "epoch": 0.14645320593468217,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9092,
+      "step": 2872
+    },
+    {
+      "epoch": 0.14650419939078754,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8676,
+      "step": 2873
+    },
+    {
+      "epoch": 0.14655519284689295,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 2874
+    },
+    {
+      "epoch": 0.14660618630299835,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8811,
+      "step": 2875
+    },
+    {
+      "epoch": 0.14665717975910372,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 2876
+    },
+    {
+      "epoch": 0.14670817321520913,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8808,
+      "step": 2877
+    },
+    {
+      "epoch": 0.1467591666713145,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 2878
+    },
+    {
+      "epoch": 0.1468101601274199,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 2879
+    },
+    {
+      "epoch": 0.14686115358352528,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14691214703963068,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 2881
+    },
+    {
+      "epoch": 0.14696314049573608,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 2882
+    },
+    {
+      "epoch": 0.14701413395184146,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 2883
+    },
+    {
+      "epoch": 0.14706512740794686,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8748,
+      "step": 2884
+    },
+    {
+      "epoch": 0.14711612086405224,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 2885
+    },
+    {
+      "epoch": 0.14716711432015764,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8528,
+      "step": 2886
+    },
+    {
+      "epoch": 0.147218107776263,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8684,
+      "step": 2887
+    },
+    {
+      "epoch": 0.14726910123236842,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8733,
+      "step": 2888
+    },
+    {
+      "epoch": 0.14732009468847382,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8803,
+      "step": 2889
+    },
+    {
+      "epoch": 0.1473710881445792,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1474220816006846,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8937,
+      "step": 2891
+    },
+    {
+      "epoch": 0.14747307505678997,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 2892
+    },
+    {
+      "epoch": 0.14752406851289537,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8731,
+      "step": 2893
+    },
+    {
+      "epoch": 0.14757506196900075,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 2894
+    },
+    {
+      "epoch": 0.14762605542510615,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 2895
+    },
+    {
+      "epoch": 0.14767704888121155,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 2896
+    },
+    {
+      "epoch": 0.14772804233731693,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8904,
+      "step": 2897
+    },
+    {
+      "epoch": 0.14777903579342233,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1478300292495277,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 2899
+    },
+    {
+      "epoch": 0.1478810227056331,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8827,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14793201616173848,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 2901
+    },
+    {
+      "epoch": 0.14798300961784389,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 2902
+    },
+    {
+      "epoch": 0.1480340030739493,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8782,
+      "step": 2903
+    },
+    {
+      "epoch": 0.14808499653005466,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.9081,
+      "step": 2904
+    },
+    {
+      "epoch": 0.14813598998616007,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.9101,
+      "step": 2905
+    },
+    {
+      "epoch": 0.14818698344226544,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8676,
+      "step": 2906
+    },
+    {
+      "epoch": 0.14823797689837084,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 2907
+    },
+    {
+      "epoch": 0.14828897035447622,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 2908
+    },
+    {
+      "epoch": 0.14833996381058162,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 2909
+    },
+    {
+      "epoch": 0.14839095726668702,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1484419507227924,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 2911
+    },
+    {
+      "epoch": 0.1484929441788978,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 2912
+    },
+    {
+      "epoch": 0.14854393763500318,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8884,
+      "step": 2913
+    },
+    {
+      "epoch": 0.14859493109110858,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8992,
+      "step": 2914
+    },
+    {
+      "epoch": 0.14864592454721395,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 2915
+    },
+    {
+      "epoch": 0.14869691800331936,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.9264,
+      "step": 2916
+    },
+    {
+      "epoch": 0.14874791145942476,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8643,
+      "step": 2917
+    },
+    {
+      "epoch": 0.14879890491553013,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 2918
+    },
+    {
+      "epoch": 0.14884989837163554,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 2919
+    },
+    {
+      "epoch": 0.1489008918277409,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1489518852838463,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8927,
+      "step": 2921
+    },
+    {
+      "epoch": 0.1490028787399517,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1490538721960571,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1491048656521625,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 2924
+    },
+    {
+      "epoch": 0.14915585910826787,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8886,
+      "step": 2925
+    },
+    {
+      "epoch": 0.14920685256437327,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 2926
+    },
+    {
+      "epoch": 0.14925784602047865,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 2927
+    },
+    {
+      "epoch": 0.14930883947658405,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 2928
+    },
+    {
+      "epoch": 0.14935983293268942,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8757,
+      "step": 2929
+    },
+    {
+      "epoch": 0.14941082638879483,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14946181984490023,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 2931
+    },
+    {
+      "epoch": 0.1495128133010056,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 2932
+    },
+    {
+      "epoch": 0.149563806757111,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.9081,
+      "step": 2933
+    },
+    {
+      "epoch": 0.14961480021321638,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8861,
+      "step": 2934
+    },
+    {
+      "epoch": 0.14966579366932178,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 2935
+    },
+    {
+      "epoch": 0.14971678712542716,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8963,
+      "step": 2936
+    },
+    {
+      "epoch": 0.14976778058153256,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 2937
+    },
+    {
+      "epoch": 0.14981877403763796,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8905,
+      "step": 2938
+    },
+    {
+      "epoch": 0.14986976749374334,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.9321,
+      "step": 2939
+    },
+    {
+      "epoch": 0.14992076094984874,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14997175440595412,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 2941
+    },
+    {
+      "epoch": 0.15002274786205952,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8924,
+      "step": 2942
+    },
+    {
+      "epoch": 0.1500737413181649,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8639,
+      "step": 2943
+    },
+    {
+      "epoch": 0.1501247347742703,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 2944
+    },
+    {
+      "epoch": 0.1501757282303757,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 2945
+    },
+    {
+      "epoch": 0.15022672168648107,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8893,
+      "step": 2946
+    },
+    {
+      "epoch": 0.15027771514258648,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 2947
+    },
+    {
+      "epoch": 0.15032870859869185,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 2948
+    },
+    {
+      "epoch": 0.15037970205479725,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8778,
+      "step": 2949
+    },
+    {
+      "epoch": 0.15043069551090263,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 2950
+    },
+    {
+      "epoch": 0.15048168896700803,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.885,
+      "step": 2951
+    },
+    {
+      "epoch": 0.15053268242311343,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 2952
+    },
+    {
+      "epoch": 0.1505836758792188,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 2953
+    },
+    {
+      "epoch": 0.1506346693353242,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8629,
+      "step": 2954
+    },
+    {
+      "epoch": 0.1506856627914296,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 2955
+    },
+    {
+      "epoch": 0.150736656247535,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8642,
+      "step": 2956
+    },
+    {
+      "epoch": 0.15078764970364036,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8701,
+      "step": 2957
+    },
+    {
+      "epoch": 0.15083864315974577,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8895,
+      "step": 2958
+    },
+    {
+      "epoch": 0.15088963661585117,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 2959
+    },
+    {
+      "epoch": 0.15094063007195654,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15099162352806195,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8838,
+      "step": 2961
+    },
+    {
+      "epoch": 0.15104261698416732,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8778,
+      "step": 2962
+    },
+    {
+      "epoch": 0.15109361044027272,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 2963
+    },
+    {
+      "epoch": 0.1511446038963781,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8694,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1511955973524835,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1512465908085889,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 2966
+    },
+    {
+      "epoch": 0.15129758426469428,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.881,
+      "step": 2967
+    },
+    {
+      "epoch": 0.15134857772079968,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8929,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15139957117690506,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8914,
+      "step": 2969
+    },
+    {
+      "epoch": 0.15145056463301046,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15150155808911583,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 2971
+    },
+    {
+      "epoch": 0.15155255154522124,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 2972
+    },
+    {
+      "epoch": 0.15160354500132664,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.9058,
+      "step": 2973
+    },
+    {
+      "epoch": 0.15165453845743201,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 2974
+    },
+    {
+      "epoch": 0.15170553191353742,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1517565253696428,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 2976
+    },
+    {
+      "epoch": 0.1518075188257482,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 2977
+    },
+    {
+      "epoch": 0.15185851228185357,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 2978
+    },
+    {
+      "epoch": 0.15190950573795897,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8756,
+      "step": 2979
+    },
+    {
+      "epoch": 0.15196049919406437,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15201149265016975,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 2981
+    },
+    {
+      "epoch": 0.15206248610627515,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 2982
+    },
+    {
+      "epoch": 0.15211347956238053,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 2983
+    },
+    {
+      "epoch": 0.15216447301848593,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8676,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1522154664745913,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8909,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1522664599306967,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 2986
+    },
+    {
+      "epoch": 0.1523174533868021,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 2987
+    },
+    {
+      "epoch": 0.15236844684290748,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 2988
+    },
+    {
+      "epoch": 0.1524194402990129,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 2989
+    },
+    {
+      "epoch": 0.15247043375511826,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15252142721122366,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 2991
+    },
+    {
+      "epoch": 0.15257242066732904,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8941,
+      "step": 2992
+    },
+    {
+      "epoch": 0.15262341412343444,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8629,
+      "step": 2993
+    },
+    {
+      "epoch": 0.15267440757953984,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 2994
+    },
+    {
+      "epoch": 0.15272540103564522,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 2995
+    },
+    {
+      "epoch": 0.15277639449175062,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 2996
+    },
+    {
+      "epoch": 0.152827387947856,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8745,
+      "step": 2997
+    },
+    {
+      "epoch": 0.1528783814039614,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 2998
+    },
+    {
+      "epoch": 0.15292937486006677,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 2999
+    },
+    {
+      "epoch": 0.15298036831617218,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15303136177227758,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8785,
+      "step": 3001
+    },
+    {
+      "epoch": 0.15308235522838295,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 3002
+    },
+    {
+      "epoch": 0.15313334868448836,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8525,
+      "step": 3003
+    },
+    {
+      "epoch": 0.15318434214059373,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8881,
+      "step": 3004
+    },
+    {
+      "epoch": 0.15323533559669913,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1532863290528045,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 3006
+    },
+    {
+      "epoch": 0.1533373225089099,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 3007
+    },
+    {
+      "epoch": 0.15338831596501531,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3008
+    },
+    {
+      "epoch": 0.1534393094211207,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 3009
+    },
+    {
+      "epoch": 0.1534903028772261,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15354129633333147,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 3011
+    },
+    {
+      "epoch": 0.15359228978943687,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8832,
+      "step": 3012
+    },
+    {
+      "epoch": 0.15364328324554224,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8902,
+      "step": 3013
+    },
+    {
+      "epoch": 0.15369427670164765,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 3014
+    },
+    {
+      "epoch": 0.15374527015775305,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 3015
+    },
+    {
+      "epoch": 0.15379626361385842,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 3016
+    },
+    {
+      "epoch": 0.15384725706996383,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 3017
+    },
+    {
+      "epoch": 0.1538982505260692,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 3018
+    },
+    {
+      "epoch": 0.1539492439821746,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 3019
+    },
+    {
+      "epoch": 0.15400023743827998,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15405123089438538,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 3021
+    },
+    {
+      "epoch": 0.15410222435049079,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 3022
+    },
+    {
+      "epoch": 0.15415321780659616,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 3023
+    },
+    {
+      "epoch": 0.15420421126270156,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 3024
+    },
+    {
+      "epoch": 0.15425520471880694,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8935,
+      "step": 3025
+    },
+    {
+      "epoch": 0.15430619817491234,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 3026
+    },
+    {
+      "epoch": 0.15435719163101772,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 3027
+    },
+    {
+      "epoch": 0.15440818508712312,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 3028
+    },
+    {
+      "epoch": 0.15445917854322852,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 3029
+    },
+    {
+      "epoch": 0.1545101719993339,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.9011,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1545611654554393,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 3031
+    },
+    {
+      "epoch": 0.15461215891154467,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8873,
+      "step": 3032
+    },
+    {
+      "epoch": 0.15466315236765008,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 3033
+    },
+    {
+      "epoch": 0.15471414582375545,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.864,
+      "step": 3034
+    },
+    {
+      "epoch": 0.15476513927986085,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 3035
+    },
+    {
+      "epoch": 0.15481613273596626,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 3036
+    },
+    {
+      "epoch": 0.15486712619207163,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 3037
+    },
+    {
+      "epoch": 0.15491811964817703,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1549691131042824,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8635,
+      "step": 3039
+    },
+    {
+      "epoch": 0.1550201065603878,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8739,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15507110001649319,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 3041
+    },
+    {
+      "epoch": 0.1551220934725986,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 3042
+    },
+    {
+      "epoch": 0.155173086928704,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 3043
+    },
+    {
+      "epoch": 0.15522408038480937,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.894,
+      "step": 3044
+    },
+    {
+      "epoch": 0.15527507384091477,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 3045
+    },
+    {
+      "epoch": 0.15532606729702014,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 3046
+    },
+    {
+      "epoch": 0.15537706075312555,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 3047
+    },
+    {
+      "epoch": 0.15542805420923092,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 3048
+    },
+    {
+      "epoch": 0.15547904766533632,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 3049
+    },
+    {
+      "epoch": 0.15553004112144173,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1555810345775471,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8848,
+      "step": 3051
+    },
+    {
+      "epoch": 0.1556320280336525,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8899,
+      "step": 3052
+    },
+    {
+      "epoch": 0.15568302148975788,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8712,
+      "step": 3053
+    },
+    {
+      "epoch": 0.15573401494586328,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 3054
+    },
+    {
+      "epoch": 0.15578500840196866,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 3055
+    },
+    {
+      "epoch": 0.15583600185807406,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8869,
+      "step": 3056
+    },
+    {
+      "epoch": 0.15588699531417946,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 3057
+    },
+    {
+      "epoch": 0.15593798877028484,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 3058
+    },
+    {
+      "epoch": 0.15598898222639024,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8816,
+      "step": 3059
+    },
+    {
+      "epoch": 0.1560399756824956,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8543,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15609096913860102,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8895,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1561419625947064,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 3062
+    },
+    {
+      "epoch": 0.1561929560508118,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 3063
+    },
+    {
+      "epoch": 0.1562439495069172,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 3064
+    },
+    {
+      "epoch": 0.15629494296302257,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 3065
+    },
+    {
+      "epoch": 0.15634593641912797,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8967,
+      "step": 3066
+    },
+    {
+      "epoch": 0.15639692987523335,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8767,
+      "step": 3067
+    },
+    {
+      "epoch": 0.15644792333133875,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 3068
+    },
+    {
+      "epoch": 0.15649891678744413,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 3069
+    },
+    {
+      "epoch": 0.15654991024354953,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8962,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15660090369965493,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8719,
+      "step": 3071
+    },
+    {
+      "epoch": 0.1566518971557603,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8751,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1567028906118657,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 3073
+    },
+    {
+      "epoch": 0.15675388406797108,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8649,
+      "step": 3074
+    },
+    {
+      "epoch": 0.15680487752407649,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 3075
+    },
+    {
+      "epoch": 0.15685587098018186,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.883,
+      "step": 3076
+    },
+    {
+      "epoch": 0.15690686443628726,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8622,
+      "step": 3077
+    },
+    {
+      "epoch": 0.15695785789239267,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 3078
+    },
+    {
+      "epoch": 0.15700885134849804,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8712,
+      "step": 3079
+    },
+    {
+      "epoch": 0.15705984480460344,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15711083826070882,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 3081
+    },
+    {
+      "epoch": 0.15716183171681422,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 3082
+    },
+    {
+      "epoch": 0.1572128251729196,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 3083
+    },
+    {
+      "epoch": 0.157263818629025,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.864,
+      "step": 3084
+    },
+    {
+      "epoch": 0.1573148120851304,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 3085
+    },
+    {
+      "epoch": 0.15736580554123578,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.885,
+      "step": 3086
+    },
+    {
+      "epoch": 0.15741679899734118,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.9035,
+      "step": 3087
+    },
+    {
+      "epoch": 0.15746779245344655,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 3088
+    },
+    {
+      "epoch": 0.15751878590955196,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8768,
+      "step": 3089
+    },
+    {
+      "epoch": 0.15756977936565733,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15762077282176273,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 3091
+    },
+    {
+      "epoch": 0.15767176627786814,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 3092
+    },
+    {
+      "epoch": 0.1577227597339735,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8743,
+      "step": 3093
+    },
+    {
+      "epoch": 0.1577737531900789,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 3094
+    },
+    {
+      "epoch": 0.1578247466461843,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8738,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1578757401022897,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 3096
+    },
+    {
+      "epoch": 0.15792673355839507,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 3097
+    },
+    {
+      "epoch": 0.15797772701450047,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8888,
+      "step": 3098
+    },
+    {
+      "epoch": 0.15802872047060587,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.9171,
+      "step": 3099
+    },
+    {
+      "epoch": 0.15807971392671125,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8654,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15813070738281665,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 3101
+    },
+    {
+      "epoch": 0.15818170083892202,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 3102
+    },
+    {
+      "epoch": 0.15823269429502743,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 3103
+    },
+    {
+      "epoch": 0.1582836877511328,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 3104
+    },
+    {
+      "epoch": 0.1583346812072382,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 3105
+    },
+    {
+      "epoch": 0.1583856746633436,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.9027,
+      "step": 3106
+    },
+    {
+      "epoch": 0.15843666811944898,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 3107
+    },
+    {
+      "epoch": 0.15848766157555438,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 3108
+    },
+    {
+      "epoch": 0.15853865503165976,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 3109
+    },
+    {
+      "epoch": 0.15858964848776516,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 3110
+    },
+    {
+      "epoch": 0.15864064194387054,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8733,
+      "step": 3111
+    },
+    {
+      "epoch": 0.15869163539997594,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 3112
+    },
+    {
+      "epoch": 0.15874262885608134,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 3113
+    },
+    {
+      "epoch": 0.15879362231218672,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8975,
+      "step": 3114
+    },
+    {
+      "epoch": 0.15884461576829212,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8957,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1588956092243975,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 3116
+    },
+    {
+      "epoch": 0.1589466026805029,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 3117
+    },
+    {
+      "epoch": 0.15899759613660827,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 3118
+    },
+    {
+      "epoch": 0.15904858959271367,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8798,
+      "step": 3119
+    },
+    {
+      "epoch": 0.15909958304881908,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8973,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15915057650492445,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8827,
+      "step": 3121
+    },
+    {
+      "epoch": 0.15920156996102985,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 3122
+    },
+    {
+      "epoch": 0.15925256341713523,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 3123
+    },
+    {
+      "epoch": 0.15930355687324063,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8731,
+      "step": 3124
+    },
+    {
+      "epoch": 0.159354550329346,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8963,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1594055437854514,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 3126
+    },
+    {
+      "epoch": 0.1594565372415568,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 3127
+    },
+    {
+      "epoch": 0.1595075306976622,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 3128
+    },
+    {
+      "epoch": 0.1595585241537676,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 3129
+    },
+    {
+      "epoch": 0.15960951760987296,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 3130
+    },
+    {
+      "epoch": 0.15966051106597837,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 3131
+    },
+    {
+      "epoch": 0.15971150452208374,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 3132
+    },
+    {
+      "epoch": 0.15976249797818914,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 3133
+    },
+    {
+      "epoch": 0.15981349143429455,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8871,
+      "step": 3134
+    },
+    {
+      "epoch": 0.15986448489039992,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 3135
+    },
+    {
+      "epoch": 0.15991547834650532,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 3136
+    },
+    {
+      "epoch": 0.1599664718026107,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 3137
+    },
+    {
+      "epoch": 0.1600174652587161,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 3138
+    },
+    {
+      "epoch": 0.16006845871482148,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.877,
+      "step": 3139
+    },
+    {
+      "epoch": 0.16011945217092688,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16017044562703228,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 3141
+    },
+    {
+      "epoch": 0.16022143908313766,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 3142
+    },
+    {
+      "epoch": 0.16027243253924306,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 3143
+    },
+    {
+      "epoch": 0.16032342599534843,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 3144
+    },
+    {
+      "epoch": 0.16037441945145384,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 3145
+    },
+    {
+      "epoch": 0.1604254129075592,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 3146
+    },
+    {
+      "epoch": 0.16047640636366461,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8679,
+      "step": 3147
+    },
+    {
+      "epoch": 0.16052739981977002,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 3148
+    },
+    {
+      "epoch": 0.1605783932758754,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 3149
+    },
+    {
+      "epoch": 0.1606293867319808,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16068038018808617,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8706,
+      "step": 3151
+    },
+    {
+      "epoch": 0.16073137364419157,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 3152
+    },
+    {
+      "epoch": 0.16078236710029695,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8877,
+      "step": 3153
+    },
+    {
+      "epoch": 0.16083336055640235,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8762,
+      "step": 3154
+    },
+    {
+      "epoch": 0.16088435401250775,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 3155
+    },
+    {
+      "epoch": 0.16093534746861313,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 3156
+    },
+    {
+      "epoch": 0.16098634092471853,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8911,
+      "step": 3157
+    },
+    {
+      "epoch": 0.1610373343808239,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8862,
+      "step": 3158
+    },
+    {
+      "epoch": 0.1610883278369293,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8889,
+      "step": 3159
+    },
+    {
+      "epoch": 0.16113932129303468,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16119031474914008,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 3161
+    },
+    {
+      "epoch": 0.1612413082052455,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8806,
+      "step": 3162
+    },
+    {
+      "epoch": 0.16129230166135086,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8795,
+      "step": 3163
+    },
+    {
+      "epoch": 0.16134329511745626,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 3164
+    },
+    {
+      "epoch": 0.16139428857356164,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 3165
+    },
+    {
+      "epoch": 0.16144528202966704,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 3166
+    },
+    {
+      "epoch": 0.16149627548577242,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 3167
+    },
+    {
+      "epoch": 0.16154726894187782,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8795,
+      "step": 3168
+    },
+    {
+      "epoch": 0.16159826239798322,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 3169
+    },
+    {
+      "epoch": 0.1616492558540886,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 3170
+    },
+    {
+      "epoch": 0.161700249310194,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8854,
+      "step": 3171
+    },
+    {
+      "epoch": 0.16175124276629937,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 3172
+    },
+    {
+      "epoch": 0.16180223622240478,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 3173
+    },
+    {
+      "epoch": 0.16185322967851015,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 3174
+    },
+    {
+      "epoch": 0.16190422313461555,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 3175
+    },
+    {
+      "epoch": 0.16195521659072096,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 3176
+    },
+    {
+      "epoch": 0.16200621004682633,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 3177
+    },
+    {
+      "epoch": 0.16205720350293173,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 3178
+    },
+    {
+      "epoch": 0.1621081969590371,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9076,
+      "step": 3179
+    },
+    {
+      "epoch": 0.1621591904151425,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 3180
+    },
+    {
+      "epoch": 0.1622101838712479,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8794,
+      "step": 3181
+    },
+    {
+      "epoch": 0.1622611773273533,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8687,
+      "step": 3182
+    },
+    {
+      "epoch": 0.1623121707834587,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 3183
+    },
+    {
+      "epoch": 0.16236316423956407,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 3184
+    },
+    {
+      "epoch": 0.16241415769566947,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 3185
+    },
+    {
+      "epoch": 0.16246515115177484,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8749,
+      "step": 3186
+    },
+    {
+      "epoch": 0.16251614460788025,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 3187
+    },
+    {
+      "epoch": 0.16256713806398562,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 3188
+    },
+    {
+      "epoch": 0.16261813152009102,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 3189
+    },
+    {
+      "epoch": 0.16266912497619643,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8794,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1627201184323018,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8843,
+      "step": 3191
+    },
+    {
+      "epoch": 0.1627711118884072,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 3192
+    },
+    {
+      "epoch": 0.16282210534451258,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 3193
+    },
+    {
+      "epoch": 0.16287309880061798,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 3194
+    },
+    {
+      "epoch": 0.16292409225672336,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 3195
+    },
+    {
+      "epoch": 0.16297508571282876,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 3196
+    },
+    {
+      "epoch": 0.16302607916893416,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 3197
+    },
+    {
+      "epoch": 0.16307707262503954,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 3198
+    },
+    {
+      "epoch": 0.16312806608114494,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 3199
+    },
+    {
+      "epoch": 0.16317905953725031,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.866,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16323005299335572,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 3201
+    },
+    {
+      "epoch": 0.1632810464494611,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8954,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1633320399055665,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8789,
+      "step": 3203
+    },
+    {
+      "epoch": 0.1633830333616719,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8828,
+      "step": 3204
+    },
+    {
+      "epoch": 0.16343402681777727,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 3205
+    },
+    {
+      "epoch": 0.16348502027388268,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 3206
+    },
+    {
+      "epoch": 0.16353601372998805,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3207
+    },
+    {
+      "epoch": 0.16358700718609345,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 3208
+    },
+    {
+      "epoch": 0.16363800064219883,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 3209
+    },
+    {
+      "epoch": 0.16368899409830423,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16373998755440963,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 3211
+    },
+    {
+      "epoch": 0.163790981010515,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 3212
+    },
+    {
+      "epoch": 0.1638419744666204,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 3213
+    },
+    {
+      "epoch": 0.16389296792272579,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 3214
+    },
+    {
+      "epoch": 0.1639439613788312,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8881,
+      "step": 3215
+    },
+    {
+      "epoch": 0.16399495483493656,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 3216
+    },
+    {
+      "epoch": 0.16404594829104197,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 3217
+    },
+    {
+      "epoch": 0.16409694174714737,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8796,
+      "step": 3218
+    },
+    {
+      "epoch": 0.16414793520325274,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8655,
+      "step": 3219
+    },
+    {
+      "epoch": 0.16419892865935815,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16424992211546352,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8928,
+      "step": 3221
+    },
+    {
+      "epoch": 0.16430091557156892,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 3222
+    },
+    {
+      "epoch": 0.1643519090276743,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8599,
+      "step": 3223
+    },
+    {
+      "epoch": 0.1644029024837797,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 3224
+    },
+    {
+      "epoch": 0.1644538959398851,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 3225
+    },
+    {
+      "epoch": 0.16450488939599048,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 3226
+    },
+    {
+      "epoch": 0.16455588285209588,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 3227
+    },
+    {
+      "epoch": 0.16460687630820126,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 3228
+    },
+    {
+      "epoch": 0.16465786976430666,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 3229
+    },
+    {
+      "epoch": 0.16470886322041203,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 3230
+    },
+    {
+      "epoch": 0.16475985667651744,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 3231
+    },
+    {
+      "epoch": 0.16481085013262284,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1648618435887282,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 3233
+    },
+    {
+      "epoch": 0.16491283704483362,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 3234
+    },
+    {
+      "epoch": 0.164963830500939,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8729,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1650148239570444,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3236
+    },
+    {
+      "epoch": 0.16506581741314977,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 3237
+    },
+    {
+      "epoch": 0.16511681086925517,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 3238
+    },
+    {
+      "epoch": 0.16516780432536057,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 3239
+    },
+    {
+      "epoch": 0.16521879778146595,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8698,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16526979123757135,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8638,
+      "step": 3241
+    },
+    {
+      "epoch": 0.16532078469367673,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 3242
+    },
+    {
+      "epoch": 0.16537177814978213,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.86,
+      "step": 3243
+    },
+    {
+      "epoch": 0.1654227716058875,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8924,
+      "step": 3244
+    },
+    {
+      "epoch": 0.1654737650619929,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 3245
+    },
+    {
+      "epoch": 0.1655247585180983,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 3246
+    },
+    {
+      "epoch": 0.16557575197420368,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 3247
+    },
+    {
+      "epoch": 0.16562674543030909,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8979,
+      "step": 3248
+    },
+    {
+      "epoch": 0.16567773888641446,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8649,
+      "step": 3249
+    },
+    {
+      "epoch": 0.16572873234251986,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16577972579862524,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 3251
+    },
+    {
+      "epoch": 0.16583071925473064,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8955,
+      "step": 3252
+    },
+    {
+      "epoch": 0.16588171271083604,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 3253
+    },
+    {
+      "epoch": 0.16593270616694142,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 3254
+    },
+    {
+      "epoch": 0.16598369962304682,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8773,
+      "step": 3255
+    },
+    {
+      "epoch": 0.1660346930791522,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.9019,
+      "step": 3256
+    },
+    {
+      "epoch": 0.1660856865352576,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 3257
+    },
+    {
+      "epoch": 0.16613667999136297,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8916,
+      "step": 3258
+    },
+    {
+      "epoch": 0.16618767344746838,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 3259
+    },
+    {
+      "epoch": 0.16623866690357378,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16628966035967915,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8843,
+      "step": 3261
+    },
+    {
+      "epoch": 0.16634065381578456,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 3262
+    },
+    {
+      "epoch": 0.16639164727188993,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8731,
+      "step": 3263
+    },
+    {
+      "epoch": 0.16644264072799533,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8847,
+      "step": 3264
+    },
+    {
+      "epoch": 0.1664936341841007,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1665446276402061,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 3266
+    },
+    {
+      "epoch": 0.1665956210963115,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 3267
+    },
+    {
+      "epoch": 0.1666466145524169,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1666976080085223,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8934,
+      "step": 3269
+    },
+    {
+      "epoch": 0.16674860146462767,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.866,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16679959492073307,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 3271
+    },
+    {
+      "epoch": 0.16685058837683844,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 3272
+    },
+    {
+      "epoch": 0.16690158183294385,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 3273
+    },
+    {
+      "epoch": 0.16695257528904925,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 3274
+    },
+    {
+      "epoch": 0.16700356874515462,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 3275
+    },
+    {
+      "epoch": 0.16705456220126003,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8963,
+      "step": 3276
+    },
+    {
+      "epoch": 0.1671055556573654,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 3277
+    },
+    {
+      "epoch": 0.1671565491134708,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 3278
+    },
+    {
+      "epoch": 0.16720754256957618,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 3279
+    },
+    {
+      "epoch": 0.16725853602568158,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8713,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16730952948178698,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 3281
+    },
+    {
+      "epoch": 0.16736052293789236,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8655,
+      "step": 3282
+    },
+    {
+      "epoch": 0.16741151639399776,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 3283
+    },
+    {
+      "epoch": 0.16746250985010314,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 3284
+    },
+    {
+      "epoch": 0.16751350330620854,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1675644967623139,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 3286
+    },
+    {
+      "epoch": 0.16761549021841932,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 3287
+    },
+    {
+      "epoch": 0.16766648367452472,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 3288
+    },
+    {
+      "epoch": 0.1677174771306301,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 3289
+    },
+    {
+      "epoch": 0.1677684705867355,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 3290
+    },
+    {
+      "epoch": 0.16781946404284087,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 3291
+    },
+    {
+      "epoch": 0.16787045749894627,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8845,
+      "step": 3292
+    },
+    {
+      "epoch": 0.16792145095505165,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 3293
+    },
+    {
+      "epoch": 0.16797244441115705,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8795,
+      "step": 3294
+    },
+    {
+      "epoch": 0.16802343786726245,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 3295
+    },
+    {
+      "epoch": 0.16807443132336783,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 3296
+    },
+    {
+      "epoch": 0.16812542477947323,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 3297
+    },
+    {
+      "epoch": 0.1681764182355786,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 3298
+    },
+    {
+      "epoch": 0.168227411691684,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8763,
+      "step": 3299
+    },
+    {
+      "epoch": 0.16827840514778938,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1683293986038948,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 3301
+    },
+    {
+      "epoch": 0.1683803920600002,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 3302
+    },
+    {
+      "epoch": 0.16843138551610556,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 3303
+    },
+    {
+      "epoch": 0.16848237897221097,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 3304
+    },
+    {
+      "epoch": 0.16853337242831634,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 3305
+    },
+    {
+      "epoch": 0.16858436588442174,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8445,
+      "step": 3306
+    },
+    {
+      "epoch": 0.16863535934052712,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8993,
+      "step": 3307
+    },
+    {
+      "epoch": 0.16868635279663252,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8787,
+      "step": 3308
+    },
+    {
+      "epoch": 0.16873734625273792,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 3309
+    },
+    {
+      "epoch": 0.1687883397088433,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1688393331649487,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8951,
+      "step": 3311
+    },
+    {
+      "epoch": 0.16889032662105408,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.895,
+      "step": 3312
+    },
+    {
+      "epoch": 0.16894132007715948,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8667,
+      "step": 3313
+    },
+    {
+      "epoch": 0.16899231353326485,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 3314
+    },
+    {
+      "epoch": 0.16904330698937026,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3315
+    },
+    {
+      "epoch": 0.16909430044547566,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8798,
+      "step": 3316
+    },
+    {
+      "epoch": 0.16914529390158103,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 3317
+    },
+    {
+      "epoch": 0.16919628735768644,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 3318
+    },
+    {
+      "epoch": 0.1692472808137918,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 3319
+    },
+    {
+      "epoch": 0.16929827426989721,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1693492677260026,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 3321
+    },
+    {
+      "epoch": 0.169400261182108,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 3322
+    },
+    {
+      "epoch": 0.1694512546382134,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8698,
+      "step": 3323
+    },
+    {
+      "epoch": 0.16950224809431877,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3324
+    },
+    {
+      "epoch": 0.16955324155042417,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.889,
+      "step": 3325
+    },
+    {
+      "epoch": 0.16960423500652955,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8737,
+      "step": 3326
+    },
+    {
+      "epoch": 0.16965522846263495,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 3327
+    },
+    {
+      "epoch": 0.16970622191874032,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8761,
+      "step": 3328
+    },
+    {
+      "epoch": 0.16975721537484573,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8818,
+      "step": 3329
+    },
+    {
+      "epoch": 0.1698082088309511,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8794,
+      "step": 3330
+    },
+    {
+      "epoch": 0.1698592022870565,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 3331
+    },
+    {
+      "epoch": 0.1699101957431619,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8846,
+      "step": 3332
+    },
+    {
+      "epoch": 0.16996118919926728,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.892,
+      "step": 3333
+    },
+    {
+      "epoch": 0.17001218265537268,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8624,
+      "step": 3334
+    },
+    {
+      "epoch": 0.17006317611147806,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8855,
+      "step": 3335
+    },
+    {
+      "epoch": 0.17011416956758346,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 3336
+    },
+    {
+      "epoch": 0.17016516302368884,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8762,
+      "step": 3337
+    },
+    {
+      "epoch": 0.17021615647979424,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 3338
+    },
+    {
+      "epoch": 0.17026714993589964,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8767,
+      "step": 3339
+    },
+    {
+      "epoch": 0.17031814339200502,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8836,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17036913684811042,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 3341
+    },
+    {
+      "epoch": 0.1704201303042158,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1704711237603212,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8684,
+      "step": 3343
+    },
+    {
+      "epoch": 0.17052211721642657,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8827,
+      "step": 3344
+    },
+    {
+      "epoch": 0.17057311067253197,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 3345
+    },
+    {
+      "epoch": 0.17062410412863738,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 3346
+    },
+    {
+      "epoch": 0.17067509758474275,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8767,
+      "step": 3347
+    },
+    {
+      "epoch": 0.17072609104084815,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8648,
+      "step": 3348
+    },
+    {
+      "epoch": 0.17077708449695353,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 3349
+    },
+    {
+      "epoch": 0.17082807795305893,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1708790714091643,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 3351
+    },
+    {
+      "epoch": 0.1709300648652697,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 3352
+    },
+    {
+      "epoch": 0.1709810583213751,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8936,
+      "step": 3353
+    },
+    {
+      "epoch": 0.1710320517774805,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.897,
+      "step": 3354
+    },
+    {
+      "epoch": 0.1710830452335859,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 3355
+    },
+    {
+      "epoch": 0.17113403868969126,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 3356
+    },
+    {
+      "epoch": 0.17118503214579667,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8808,
+      "step": 3357
+    },
+    {
+      "epoch": 0.17123602560190204,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3358
+    },
+    {
+      "epoch": 0.17128701905800744,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8667,
+      "step": 3359
+    },
+    {
+      "epoch": 0.17133801251411285,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9097,
+      "step": 3360
+    },
+    {
+      "epoch": 0.17138900597021822,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 3361
+    },
+    {
+      "epoch": 0.17143999942632362,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8712,
+      "step": 3362
+    },
+    {
+      "epoch": 0.171490992882429,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 3363
+    },
+    {
+      "epoch": 0.1715419863385344,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 3364
+    },
+    {
+      "epoch": 0.17159297979463978,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 3365
+    },
+    {
+      "epoch": 0.17164397325074518,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 3366
+    },
+    {
+      "epoch": 0.17169496670685058,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 3367
+    },
+    {
+      "epoch": 0.17174596016295596,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 3368
+    },
+    {
+      "epoch": 0.17179695361906136,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 3369
+    },
+    {
+      "epoch": 0.17184794707516673,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 3370
+    },
+    {
+      "epoch": 0.17189894053127214,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8958,
+      "step": 3371
+    },
+    {
+      "epoch": 0.1719499339873775,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8746,
+      "step": 3372
+    },
+    {
+      "epoch": 0.17200092744348291,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 3373
+    },
+    {
+      "epoch": 0.17205192089958832,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.883,
+      "step": 3374
+    },
+    {
+      "epoch": 0.1721029143556937,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 3375
+    },
+    {
+      "epoch": 0.1721539078117991,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 3376
+    },
+    {
+      "epoch": 0.17220490126790447,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8934,
+      "step": 3377
+    },
+    {
+      "epoch": 0.17225589472400987,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3378
+    },
+    {
+      "epoch": 0.17230688818011525,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8886,
+      "step": 3379
+    },
+    {
+      "epoch": 0.17235788163622065,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.9048,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17240887509232605,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 3381
+    },
+    {
+      "epoch": 0.17245986854843143,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9246,
+      "step": 3382
+    },
+    {
+      "epoch": 0.17251086200453683,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 3383
+    },
+    {
+      "epoch": 0.1725618554606422,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 3384
+    },
+    {
+      "epoch": 0.1726128489167476,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 3385
+    },
+    {
+      "epoch": 0.17266384237285298,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.002,
+      "loss": 2.8655,
+      "step": 3386
+    },
+    {
+      "epoch": 0.17271483582895839,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 3387
+    },
+    {
+      "epoch": 0.1727658292850638,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 3388
+    },
+    {
+      "epoch": 0.17281682274116916,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 3389
+    },
+    {
+      "epoch": 0.17286781619727457,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17291880965337994,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8759,
+      "step": 3391
+    },
+    {
+      "epoch": 0.17296980310948534,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 3392
+    },
+    {
+      "epoch": 0.17302079656559072,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 3393
+    },
+    {
+      "epoch": 0.17307179002169612,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 3394
+    },
+    {
+      "epoch": 0.17312278347780152,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8769,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1731737769339069,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 3396
+    },
+    {
+      "epoch": 0.1732247703900123,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3397
+    },
+    {
+      "epoch": 0.17327576384611768,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3398
+    },
+    {
+      "epoch": 0.17332675730222308,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8846,
+      "step": 3399
+    },
+    {
+      "epoch": 0.17337775075832845,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17342874421443386,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 3401
+    },
+    {
+      "epoch": 0.17347973767053926,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 3402
+    },
+    {
+      "epoch": 0.17353073112664463,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8738,
+      "step": 3403
+    },
+    {
+      "epoch": 0.17358172458275004,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 3404
+    },
+    {
+      "epoch": 0.1736327180388554,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 3405
+    },
+    {
+      "epoch": 0.1736837114949608,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 3406
+    },
+    {
+      "epoch": 0.1737347049510662,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8817,
+      "step": 3407
+    },
+    {
+      "epoch": 0.1737856984071716,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 3408
+    },
+    {
+      "epoch": 0.173836691863277,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8923,
+      "step": 3409
+    },
+    {
+      "epoch": 0.17388768531938237,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17393867877548777,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 3411
+    },
+    {
+      "epoch": 0.17398967223159315,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8877,
+      "step": 3412
+    },
+    {
+      "epoch": 0.17404066568769855,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 3413
+    },
+    {
+      "epoch": 0.17409165914380392,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 3414
+    },
+    {
+      "epoch": 0.17414265259990933,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 3415
+    },
+    {
+      "epoch": 0.17419364605601473,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 3416
+    },
+    {
+      "epoch": 0.1742446395121201,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8792,
+      "step": 3417
+    },
+    {
+      "epoch": 0.1742956329682255,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 3418
+    },
+    {
+      "epoch": 0.17434662642433088,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 3419
+    },
+    {
+      "epoch": 0.17439761988043628,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17444861333654166,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 3421
+    },
+    {
+      "epoch": 0.17449960679264706,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 3422
+    },
+    {
+      "epoch": 0.17455060024875246,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 3423
+    },
+    {
+      "epoch": 0.17460159370485784,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8757,
+      "step": 3424
+    },
+    {
+      "epoch": 0.17465258716096324,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 3425
+    },
+    {
+      "epoch": 0.17470358061706862,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 3426
+    },
+    {
+      "epoch": 0.17475457407317402,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 3427
+    },
+    {
+      "epoch": 0.1748055675292794,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1748565609853848,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8931,
+      "step": 3429
+    },
+    {
+      "epoch": 0.1749075544414902,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 3430
+    },
+    {
+      "epoch": 0.17495854789759557,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 3431
+    },
+    {
+      "epoch": 0.17500954135370098,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 3432
+    },
+    {
+      "epoch": 0.17506053480980635,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.87,
+      "step": 3433
+    },
+    {
+      "epoch": 0.17511152826591175,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 3434
+    },
+    {
+      "epoch": 0.17516252172201713,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8836,
+      "step": 3435
+    },
+    {
+      "epoch": 0.17521351517812253,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8713,
+      "step": 3436
+    },
+    {
+      "epoch": 0.17526450863422793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 3437
+    },
+    {
+      "epoch": 0.1753155020903333,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8684,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1753664955464387,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 3439
+    },
+    {
+      "epoch": 0.17541748900254409,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1754684824586495,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 3441
+    },
+    {
+      "epoch": 0.17551947591475486,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8884,
+      "step": 3442
+    },
+    {
+      "epoch": 0.17557046937086027,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 3443
+    },
+    {
+      "epoch": 0.17562146282696567,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 3444
+    },
+    {
+      "epoch": 0.17567245628307104,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 3445
+    },
+    {
+      "epoch": 0.17572344973917645,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 3446
+    },
+    {
+      "epoch": 0.17577444319528182,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 3447
+    },
+    {
+      "epoch": 0.17582543665138722,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 3448
+    },
+    {
+      "epoch": 0.1758764301074926,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 3449
+    },
+    {
+      "epoch": 0.175927423563598,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8694,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1759784170197034,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8725,
+      "step": 3451
+    },
+    {
+      "epoch": 0.17602941047580878,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 3452
+    },
+    {
+      "epoch": 0.17608040393191418,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8808,
+      "step": 3453
+    },
+    {
+      "epoch": 0.17613139738801956,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8873,
+      "step": 3454
+    },
+    {
+      "epoch": 0.17618239084412496,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 3455
+    },
+    {
+      "epoch": 0.17623338430023033,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8876,
+      "step": 3456
+    },
+    {
+      "epoch": 0.17628437775633574,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3457
+    },
+    {
+      "epoch": 0.17633537121244114,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 3458
+    },
+    {
+      "epoch": 0.1763863646685465,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 3459
+    },
+    {
+      "epoch": 0.17643735812465192,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8856,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1764883515807573,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 3461
+    },
+    {
+      "epoch": 0.1765393450368627,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8642,
+      "step": 3462
+    },
+    {
+      "epoch": 0.17659033849296807,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.858,
+      "step": 3463
+    },
+    {
+      "epoch": 0.17664133194907347,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 3464
+    },
+    {
+      "epoch": 0.17669232540517887,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 3465
+    },
+    {
+      "epoch": 0.17674331886128425,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 3466
+    },
+    {
+      "epoch": 0.17679431231738965,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 3467
+    },
+    {
+      "epoch": 0.17684530577349503,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 3468
+    },
+    {
+      "epoch": 0.17689629922960043,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 3469
+    },
+    {
+      "epoch": 0.1769472926857058,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8635,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1769982861418112,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8712,
+      "step": 3471
+    },
+    {
+      "epoch": 0.1770492795979166,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8694,
+      "step": 3472
+    },
+    {
+      "epoch": 0.17710027305402198,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 3473
+    },
+    {
+      "epoch": 0.1771512665101274,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 3474
+    },
+    {
+      "epoch": 0.17720225996623276,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8815,
+      "step": 3475
+    },
+    {
+      "epoch": 0.17725325342233816,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8864,
+      "step": 3476
+    },
+    {
+      "epoch": 0.17730424687844354,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8865,
+      "step": 3477
+    },
+    {
+      "epoch": 0.17735524033454894,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 3478
+    },
+    {
+      "epoch": 0.17740623379065434,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8922,
+      "step": 3479
+    },
+    {
+      "epoch": 0.17745722724675972,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8801,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17750822070286512,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 3481
+    },
+    {
+      "epoch": 0.1775592141589705,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 3482
+    },
+    {
+      "epoch": 0.1776102076150759,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 3483
+    },
+    {
+      "epoch": 0.17766120107118127,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.883,
+      "step": 3484
+    },
+    {
+      "epoch": 0.17771219452728668,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 3485
+    },
+    {
+      "epoch": 0.17776318798339208,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 3486
+    },
+    {
+      "epoch": 0.17781418143949745,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 3487
+    },
+    {
+      "epoch": 0.17786517489560286,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8716,
+      "step": 3488
+    },
+    {
+      "epoch": 0.17791616835170823,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 3489
+    },
+    {
+      "epoch": 0.17796716180781363,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 3490
+    },
+    {
+      "epoch": 0.178018155263919,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 3491
+    },
+    {
+      "epoch": 0.1780691487200244,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 3492
+    },
+    {
+      "epoch": 0.17812014217612981,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 3493
+    },
+    {
+      "epoch": 0.1781711356322352,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 3494
+    },
+    {
+      "epoch": 0.1782221290883406,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 3495
+    },
+    {
+      "epoch": 0.17827312254444597,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 3496
+    },
+    {
+      "epoch": 0.17832411600055137,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 3497
+    },
+    {
+      "epoch": 0.17837510945665674,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 3498
+    },
+    {
+      "epoch": 0.17842610291276215,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 3499
+    },
+    {
+      "epoch": 0.17847709636886755,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8698,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17852808982497292,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 3501
+    },
+    {
+      "epoch": 0.17857908328107833,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8826,
+      "step": 3502
+    },
+    {
+      "epoch": 0.1786300767371837,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 3503
+    },
+    {
+      "epoch": 0.1786810701932891,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.894,
+      "step": 3504
+    },
+    {
+      "epoch": 0.17873206364939448,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 3505
+    },
+    {
+      "epoch": 0.17878305710549988,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 3506
+    },
+    {
+      "epoch": 0.17883405056160528,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 3507
+    },
+    {
+      "epoch": 0.17888504401771066,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 3508
+    },
+    {
+      "epoch": 0.17893603747381606,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 3509
+    },
+    {
+      "epoch": 0.17898703092992144,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17903802438602684,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 3511
+    },
+    {
+      "epoch": 0.17908901784213221,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8914,
+      "step": 3512
+    },
+    {
+      "epoch": 0.17914001129823762,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 3513
+    },
+    {
+      "epoch": 0.17919100475434302,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 3514
+    },
+    {
+      "epoch": 0.1792419982104484,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1792929916665538,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8707,
+      "step": 3516
+    },
+    {
+      "epoch": 0.17934398512265917,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 3517
+    },
+    {
+      "epoch": 0.17939497857876457,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8719,
+      "step": 3518
+    },
+    {
+      "epoch": 0.17944597203486995,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 3519
+    },
+    {
+      "epoch": 0.17949696549097535,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17954795894708075,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 3521
+    },
+    {
+      "epoch": 0.17959895240318613,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8707,
+      "step": 3522
+    },
+    {
+      "epoch": 0.17964994585929153,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 3523
+    },
+    {
+      "epoch": 0.1797009393153969,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 3524
+    },
+    {
+      "epoch": 0.1797519327715023,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 3525
+    },
+    {
+      "epoch": 0.17980292622760768,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 3526
+    },
+    {
+      "epoch": 0.1798539196837131,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8739,
+      "step": 3527
+    },
+    {
+      "epoch": 0.1799049131398185,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8945,
+      "step": 3528
+    },
+    {
+      "epoch": 0.17995590659592386,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 3529
+    },
+    {
+      "epoch": 0.18000690005202927,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18005789350813464,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8808,
+      "step": 3531
+    },
+    {
+      "epoch": 0.18010888696424004,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8783,
+      "step": 3532
+    },
+    {
+      "epoch": 0.18015988042034542,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8787,
+      "step": 3533
+    },
+    {
+      "epoch": 0.18021087387645082,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8822,
+      "step": 3534
+    },
+    {
+      "epoch": 0.18026186733255622,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1803128607886616,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8654,
+      "step": 3536
+    },
+    {
+      "epoch": 0.180363854244767,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.864,
+      "step": 3537
+    },
+    {
+      "epoch": 0.18041484770087238,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 3538
+    },
+    {
+      "epoch": 0.18046584115697778,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 3539
+    },
+    {
+      "epoch": 0.18051683461308315,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18056782806918856,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 3541
+    },
+    {
+      "epoch": 0.18061882152529396,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 3542
+    },
+    {
+      "epoch": 0.18066981498139933,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 3543
+    },
+    {
+      "epoch": 0.18072080843750474,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 3544
+    },
+    {
+      "epoch": 0.1807718018936101,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8903,
+      "step": 3545
+    },
+    {
+      "epoch": 0.18082279534971551,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 3546
+    },
+    {
+      "epoch": 0.1808737888058209,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 3547
+    },
+    {
+      "epoch": 0.1809247822619263,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8649,
+      "step": 3548
+    },
+    {
+      "epoch": 0.1809757757180317,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8849,
+      "step": 3549
+    },
+    {
+      "epoch": 0.18102676917413707,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18107776263024247,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 3551
+    },
+    {
+      "epoch": 0.18112875608634785,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.9008,
+      "step": 3552
+    },
+    {
+      "epoch": 0.18117974954245325,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 3553
+    },
+    {
+      "epoch": 0.18123074299855862,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8787,
+      "step": 3554
+    },
+    {
+      "epoch": 0.18128173645466403,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 3555
+    },
+    {
+      "epoch": 0.18133272991076943,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3556
+    },
+    {
+      "epoch": 0.1813837233668748,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 3557
+    },
+    {
+      "epoch": 0.1814347168229802,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 3558
+    },
+    {
+      "epoch": 0.18148571027908558,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 3559
+    },
+    {
+      "epoch": 0.18153670373519099,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8956,
+      "step": 3560
+    },
+    {
+      "epoch": 0.18158769719129636,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8833,
+      "step": 3561
+    },
+    {
+      "epoch": 0.18163869064740176,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8727,
+      "step": 3562
+    },
+    {
+      "epoch": 0.18168968410350717,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 3563
+    },
+    {
+      "epoch": 0.18174067755961254,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.882,
+      "step": 3564
+    },
+    {
+      "epoch": 0.18179167101571794,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 3565
+    },
+    {
+      "epoch": 0.18184266447182332,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 3566
+    },
+    {
+      "epoch": 0.18189365792792872,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 3567
+    },
+    {
+      "epoch": 0.1819446513840341,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 3568
+    },
+    {
+      "epoch": 0.1819956448401395,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 3569
+    },
+    {
+      "epoch": 0.1820466382962449,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8815,
+      "step": 3570
+    },
+    {
+      "epoch": 0.18209763175235028,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 3571
+    },
+    {
+      "epoch": 0.18214862520845568,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 3572
+    },
+    {
+      "epoch": 0.18219961866456105,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8887,
+      "step": 3573
+    },
+    {
+      "epoch": 0.18225061212066646,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8667,
+      "step": 3574
+    },
+    {
+      "epoch": 0.18230160557677183,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 3575
+    },
+    {
+      "epoch": 0.18235259903287723,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 3576
+    },
+    {
+      "epoch": 0.18240359248898264,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8782,
+      "step": 3577
+    },
+    {
+      "epoch": 0.182454585945088,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8813,
+      "step": 3578
+    },
+    {
+      "epoch": 0.1825055794011934,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8974,
+      "step": 3579
+    },
+    {
+      "epoch": 0.1825565728572988,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.888,
+      "step": 3580
+    },
+    {
+      "epoch": 0.1826075663134042,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 3581
+    },
+    {
+      "epoch": 0.18265855976950957,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8813,
+      "step": 3582
+    },
+    {
+      "epoch": 0.18270955322561497,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 3583
+    },
+    {
+      "epoch": 0.18276054668172037,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8934,
+      "step": 3584
+    },
+    {
+      "epoch": 0.18281154013782575,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8689,
+      "step": 3585
+    },
+    {
+      "epoch": 0.18286253359393115,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8869,
+      "step": 3586
+    },
+    {
+      "epoch": 0.18291352705003652,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 3587
+    },
+    {
+      "epoch": 0.18296452050614193,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 3588
+    },
+    {
+      "epoch": 0.1830155139622473,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 3589
+    },
+    {
+      "epoch": 0.1830665074183527,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8643,
+      "step": 3590
+    },
+    {
+      "epoch": 0.1831175008744581,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8756,
+      "step": 3591
+    },
+    {
+      "epoch": 0.18316849433056348,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8798,
+      "step": 3592
+    },
+    {
+      "epoch": 0.18321948778666888,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 3593
+    },
+    {
+      "epoch": 0.18327048124277426,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 3594
+    },
+    {
+      "epoch": 0.18332147469887966,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3595
+    },
+    {
+      "epoch": 0.18337246815498504,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 3596
+    },
+    {
+      "epoch": 0.18342346161109044,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 3597
+    },
+    {
+      "epoch": 0.18347445506719584,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 3598
+    },
+    {
+      "epoch": 0.18352544852330122,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 3599
+    },
+    {
+      "epoch": 0.18357644197940662,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8792,
+      "step": 3600
+    },
+    {
+      "epoch": 0.183627435435512,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 3601
+    },
+    {
+      "epoch": 0.1836784288916174,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 3602
+    },
+    {
+      "epoch": 0.18372942234772277,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8841,
+      "step": 3603
+    },
+    {
+      "epoch": 0.18378041580382817,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 3604
+    },
+    {
+      "epoch": 0.18383140925993358,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8748,
+      "step": 3605
+    },
+    {
+      "epoch": 0.18388240271603895,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8818,
+      "step": 3606
+    },
+    {
+      "epoch": 0.18393339617214435,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 3607
+    },
+    {
+      "epoch": 0.18398438962824973,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 3608
+    },
+    {
+      "epoch": 0.18403538308435513,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8622,
+      "step": 3609
+    },
+    {
+      "epoch": 0.1840863765404605,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8803,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1841373699965659,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8853,
+      "step": 3611
+    },
+    {
+      "epoch": 0.1841883634526713,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8771,
+      "step": 3612
+    },
+    {
+      "epoch": 0.18423935690877669,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8737,
+      "step": 3613
+    },
+    {
+      "epoch": 0.1842903503648821,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 3614
+    },
+    {
+      "epoch": 0.18434134382098746,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8746,
+      "step": 3615
+    },
+    {
+      "epoch": 0.18439233727709287,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3616
+    },
+    {
+      "epoch": 0.18444333073319824,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 3617
+    },
+    {
+      "epoch": 0.18449432418930364,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 3618
+    },
+    {
+      "epoch": 0.18454531764540905,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8746,
+      "step": 3619
+    },
+    {
+      "epoch": 0.18459631110151442,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18464730455761982,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 3621
+    },
+    {
+      "epoch": 0.1846982980137252,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8815,
+      "step": 3622
+    },
+    {
+      "epoch": 0.1847492914698306,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 3623
+    },
+    {
+      "epoch": 0.18480028492593598,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 3624
+    },
+    {
+      "epoch": 0.18485127838204138,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8655,
+      "step": 3625
+    },
+    {
+      "epoch": 0.18490227183814678,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 3626
+    },
+    {
+      "epoch": 0.18495326529425216,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 3627
+    },
+    {
+      "epoch": 0.18500425875035756,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8623,
+      "step": 3628
+    },
+    {
+      "epoch": 0.18505525220646293,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 3629
+    },
+    {
+      "epoch": 0.18510624566256834,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1851572391186737,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8878,
+      "step": 3631
+    },
+    {
+      "epoch": 0.1852082325747791,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8745,
+      "step": 3632
+    },
+    {
+      "epoch": 0.18525922603088452,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 3633
+    },
+    {
+      "epoch": 0.1853102194869899,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.86,
+      "step": 3634
+    },
+    {
+      "epoch": 0.1853612129430953,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 3635
+    },
+    {
+      "epoch": 0.18541220639920067,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 3636
+    },
+    {
+      "epoch": 0.18546319985530607,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 3637
+    },
+    {
+      "epoch": 0.18551419331141145,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 3638
+    },
+    {
+      "epoch": 0.18556518676751685,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 3639
+    },
+    {
+      "epoch": 0.18561618022362225,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 3640
+    },
+    {
+      "epoch": 0.18566717367972763,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8724,
+      "step": 3641
+    },
+    {
+      "epoch": 0.18571816713583303,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8905,
+      "step": 3642
+    },
+    {
+      "epoch": 0.1857691605919384,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 3643
+    },
+    {
+      "epoch": 0.1858201540480438,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 3644
+    },
+    {
+      "epoch": 0.18587114750414918,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 3645
+    },
+    {
+      "epoch": 0.18592214096025458,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 3646
+    },
+    {
+      "epoch": 0.18597313441636,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.891,
+      "step": 3647
+    },
+    {
+      "epoch": 0.18602412787246536,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 3648
+    },
+    {
+      "epoch": 0.18607512132857076,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 3649
+    },
+    {
+      "epoch": 0.18612611478467614,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 3650
+    },
+    {
+      "epoch": 0.18617710824078154,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 3651
+    },
+    {
+      "epoch": 0.18622810169688692,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 3652
+    },
+    {
+      "epoch": 0.18627909515299232,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8716,
+      "step": 3653
+    },
+    {
+      "epoch": 0.18633008860909772,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 3654
+    },
+    {
+      "epoch": 0.1863810820652031,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 3655
+    },
+    {
+      "epoch": 0.1864320755213085,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 3656
+    },
+    {
+      "epoch": 0.18648306897741387,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 3657
+    },
+    {
+      "epoch": 0.18653406243351928,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 3658
+    },
+    {
+      "epoch": 0.18658505588962465,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 3659
+    },
+    {
+      "epoch": 0.18663604934573005,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 3660
+    },
+    {
+      "epoch": 0.18668704280183546,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.888,
+      "step": 3661
+    },
+    {
+      "epoch": 0.18673803625794083,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 3662
+    },
+    {
+      "epoch": 0.18678902971404623,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 3663
+    },
+    {
+      "epoch": 0.1868400231701516,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 3664
+    },
+    {
+      "epoch": 0.186891016626257,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 3665
+    },
+    {
+      "epoch": 0.1869420100823624,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 3666
+    },
+    {
+      "epoch": 0.1869930035384678,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 3667
+    },
+    {
+      "epoch": 0.1870439969945732,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8816,
+      "step": 3668
+    },
+    {
+      "epoch": 0.18709499045067857,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 3669
+    },
+    {
+      "epoch": 0.18714598390678397,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18719697736288934,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 3671
+    },
+    {
+      "epoch": 0.18724797081899475,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 3672
+    },
+    {
+      "epoch": 0.18729896427510012,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.882,
+      "step": 3673
+    },
+    {
+      "epoch": 0.18734995773120552,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 3674
+    },
+    {
+      "epoch": 0.18740095118731093,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8607,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1874519446434163,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.9089,
+      "step": 3676
+    },
+    {
+      "epoch": 0.1875029380995217,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 3677
+    },
+    {
+      "epoch": 0.18755393155562708,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 3678
+    },
+    {
+      "epoch": 0.18760492501173248,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8912,
+      "step": 3679
+    },
+    {
+      "epoch": 0.18765591846783786,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 3680
+    },
+    {
+      "epoch": 0.18770691192394326,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 3681
+    },
+    {
+      "epoch": 0.18775790538004866,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 3682
+    },
+    {
+      "epoch": 0.18780889883615404,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 3683
+    },
+    {
+      "epoch": 0.18785989229225944,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 3684
+    },
+    {
+      "epoch": 0.18791088574836481,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 3685
+    },
+    {
+      "epoch": 0.18796187920447022,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 3686
+    },
+    {
+      "epoch": 0.1880128726605756,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 3687
+    },
+    {
+      "epoch": 0.188063866116681,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 3688
+    },
+    {
+      "epoch": 0.1881148595727864,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 3689
+    },
+    {
+      "epoch": 0.18816585302889177,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18821684648499717,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 3691
+    },
+    {
+      "epoch": 0.18826783994110255,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8919,
+      "step": 3692
+    },
+    {
+      "epoch": 0.18831883339720795,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8897,
+      "step": 3693
+    },
+    {
+      "epoch": 0.18836982685331333,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 3694
+    },
+    {
+      "epoch": 0.18842082030941873,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 3695
+    },
+    {
+      "epoch": 0.18847181376552413,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 3696
+    },
+    {
+      "epoch": 0.1885228072216295,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.874,
+      "step": 3697
+    },
+    {
+      "epoch": 0.1885738006777349,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 3698
+    },
+    {
+      "epoch": 0.18862479413384028,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 3699
+    },
+    {
+      "epoch": 0.1886757875899457,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18872678104605106,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.87,
+      "step": 3701
+    },
+    {
+      "epoch": 0.18877777450215646,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 3702
+    },
+    {
+      "epoch": 0.18882876795826187,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 3703
+    },
+    {
+      "epoch": 0.18887976141436724,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 3704
+    },
+    {
+      "epoch": 0.18893075487047264,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 3705
+    },
+    {
+      "epoch": 0.18898174832657802,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 3706
+    },
+    {
+      "epoch": 0.18903274178268342,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.875,
+      "step": 3707
+    },
+    {
+      "epoch": 0.1890837352387888,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 3708
+    },
+    {
+      "epoch": 0.1891347286948942,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 3709
+    },
+    {
+      "epoch": 0.1891857221509996,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8681,
+      "step": 3710
+    },
+    {
+      "epoch": 0.18923671560710498,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8639,
+      "step": 3711
+    },
+    {
+      "epoch": 0.18928770906321038,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 3712
+    },
+    {
+      "epoch": 0.18933870251931575,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 3713
+    },
+    {
+      "epoch": 0.18938969597542116,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 3714
+    },
+    {
+      "epoch": 0.18944068943152653,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 3715
+    },
+    {
+      "epoch": 0.18949168288763193,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 3716
+    },
+    {
+      "epoch": 0.18954267634373734,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 3717
+    },
+    {
+      "epoch": 0.1895936697998427,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.877,
+      "step": 3718
+    },
+    {
+      "epoch": 0.18964466325594811,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 3719
+    },
+    {
+      "epoch": 0.1896956567120535,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1897466501681589,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8805,
+      "step": 3721
+    },
+    {
+      "epoch": 0.18979764362426427,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 3722
+    },
+    {
+      "epoch": 0.18984863708036967,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 3723
+    },
+    {
+      "epoch": 0.18989963053647507,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 3724
+    },
+    {
+      "epoch": 0.18995062399258045,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 3725
+    },
+    {
+      "epoch": 0.19000161744868585,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 3726
+    },
+    {
+      "epoch": 0.19005261090479122,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 3727
+    },
+    {
+      "epoch": 0.19010360436089663,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 3728
+    },
+    {
+      "epoch": 0.190154597817002,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 3729
+    },
+    {
+      "epoch": 0.1902055912731074,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8673,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1902565847292128,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 3731
+    },
+    {
+      "epoch": 0.19030757818531818,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 3732
+    },
+    {
+      "epoch": 0.19035857164142359,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8939,
+      "step": 3733
+    },
+    {
+      "epoch": 0.19040956509752896,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 3734
+    },
+    {
+      "epoch": 0.19046055855363436,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 3735
+    },
+    {
+      "epoch": 0.19051155200973974,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 3736
+    },
+    {
+      "epoch": 0.19056254546584514,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 3737
+    },
+    {
+      "epoch": 0.19061353892195054,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 3738
+    },
+    {
+      "epoch": 0.19066453237805592,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 3739
+    },
+    {
+      "epoch": 0.19071552583416132,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8553,
+      "step": 3740
+    },
+    {
+      "epoch": 0.1907665192902667,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 3741
+    },
+    {
+      "epoch": 0.1908175127463721,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 3742
+    },
+    {
+      "epoch": 0.19086850620247747,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 3743
+    },
+    {
+      "epoch": 0.19091949965858288,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 3744
+    },
+    {
+      "epoch": 0.19097049311468828,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 3745
+    },
+    {
+      "epoch": 0.19102148657079365,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 3746
+    },
+    {
+      "epoch": 0.19107248002689906,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3747
+    },
+    {
+      "epoch": 0.19112347348300443,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 3748
+    },
+    {
+      "epoch": 0.19117446693910983,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 3749
+    },
+    {
+      "epoch": 0.1912254603952152,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8768,
+      "step": 3750
+    },
+    {
+      "epoch": 0.1912764538513206,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 3751
+    },
+    {
+      "epoch": 0.191327447307426,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 3752
+    },
+    {
+      "epoch": 0.1913784407635314,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 3753
+    },
+    {
+      "epoch": 0.1914294342196368,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 3754
+    },
+    {
+      "epoch": 0.19148042767574217,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 3755
+    },
+    {
+      "epoch": 0.19153142113184757,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8377,
+      "step": 3756
+    },
+    {
+      "epoch": 0.19158241458795294,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 3757
+    },
+    {
+      "epoch": 0.19163340804405835,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 3758
+    },
+    {
+      "epoch": 0.19168440150016375,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8743,
+      "step": 3759
+    },
+    {
+      "epoch": 0.19173539495626912,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19178638841237453,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8648,
+      "step": 3761
+    },
+    {
+      "epoch": 0.1918373818684799,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 3762
+    },
+    {
+      "epoch": 0.1918883753245853,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 3763
+    },
+    {
+      "epoch": 0.19193936878069068,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 3764
+    },
+    {
+      "epoch": 0.19199036223679608,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 3765
+    },
+    {
+      "epoch": 0.19204135569290148,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 3766
+    },
+    {
+      "epoch": 0.19209234914900686,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8656,
+      "step": 3767
+    },
+    {
+      "epoch": 0.19214334260511226,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 3768
+    },
+    {
+      "epoch": 0.19219433606121764,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 3769
+    },
+    {
+      "epoch": 0.19224532951732304,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 3770
+    },
+    {
+      "epoch": 0.1922963229734284,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 3771
+    },
+    {
+      "epoch": 0.19234731642953382,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8854,
+      "step": 3772
+    },
+    {
+      "epoch": 0.19239830988563922,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8891,
+      "step": 3773
+    },
+    {
+      "epoch": 0.1924493033417446,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 3774
+    },
+    {
+      "epoch": 0.19250029679785,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 3775
+    },
+    {
+      "epoch": 0.19255129025395537,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 3776
+    },
+    {
+      "epoch": 0.19260228371006077,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8923,
+      "step": 3777
+    },
+    {
+      "epoch": 0.19265327716616615,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 3778
+    },
+    {
+      "epoch": 0.19270427062227155,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8743,
+      "step": 3779
+    },
+    {
+      "epoch": 0.19275526407837695,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8785,
+      "step": 3780
+    },
+    {
+      "epoch": 0.19280625753448233,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 3781
+    },
+    {
+      "epoch": 0.19285725099058773,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.9003,
+      "step": 3782
+    },
+    {
+      "epoch": 0.1929082444466931,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8738,
+      "step": 3783
+    },
+    {
+      "epoch": 0.1929592379027985,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8737,
+      "step": 3784
+    },
+    {
+      "epoch": 0.19301023135890388,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 3785
+    },
+    {
+      "epoch": 0.19306122481500929,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3786
+    },
+    {
+      "epoch": 0.1931122182711147,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8827,
+      "step": 3787
+    },
+    {
+      "epoch": 0.19316321172722006,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 3788
+    },
+    {
+      "epoch": 0.19321420518332547,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8969,
+      "step": 3789
+    },
+    {
+      "epoch": 0.19326519863943084,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8858,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19331619209553624,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 3791
+    },
+    {
+      "epoch": 0.19336718555164162,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 3792
+    },
+    {
+      "epoch": 0.19341817900774702,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3793
+    },
+    {
+      "epoch": 0.19346917246385242,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 3794
+    },
+    {
+      "epoch": 0.1935201659199578,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1935711593760632,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8957,
+      "step": 3796
+    },
+    {
+      "epoch": 0.19362215283216858,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8479,
+      "step": 3797
+    },
+    {
+      "epoch": 0.19367314628827398,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 3798
+    },
+    {
+      "epoch": 0.19372413974437935,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 3799
+    },
+    {
+      "epoch": 0.19377513320048476,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19382612665659016,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 3801
+    },
+    {
+      "epoch": 0.19387712011269553,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 3802
+    },
+    {
+      "epoch": 0.19392811356880094,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 3803
+    },
+    {
+      "epoch": 0.1939791070249063,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 3804
+    },
+    {
+      "epoch": 0.1940301004810117,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 3805
+    },
+    {
+      "epoch": 0.1940810939371171,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 3806
+    },
+    {
+      "epoch": 0.1941320873932225,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.9052,
+      "step": 3807
+    },
+    {
+      "epoch": 0.1941830808493279,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8687,
+      "step": 3808
+    },
+    {
+      "epoch": 0.19423407430543327,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 3809
+    },
+    {
+      "epoch": 0.19428506776153867,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8814,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19433606121764405,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 3811
+    },
+    {
+      "epoch": 0.19438705467374945,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 3812
+    },
+    {
+      "epoch": 0.19443804812985482,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 3813
+    },
+    {
+      "epoch": 0.19448904158596023,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8719,
+      "step": 3814
+    },
+    {
+      "epoch": 0.19454003504206563,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8525,
+      "step": 3815
+    },
+    {
+      "epoch": 0.194591028498171,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 3816
+    },
+    {
+      "epoch": 0.1946420219542764,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 3817
+    },
+    {
+      "epoch": 0.19469301541038178,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 3818
+    },
+    {
+      "epoch": 0.19474400886648718,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 3819
+    },
+    {
+      "epoch": 0.19479500232259256,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19484599577869796,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 3821
+    },
+    {
+      "epoch": 0.19489698923480336,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 3822
+    },
+    {
+      "epoch": 0.19494798269090874,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 3823
+    },
+    {
+      "epoch": 0.19499897614701414,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 3824
+    },
+    {
+      "epoch": 0.19504996960311952,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8673,
+      "step": 3825
+    },
+    {
+      "epoch": 0.19510096305922492,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 3826
+    },
+    {
+      "epoch": 0.1951519565153303,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 3827
+    },
+    {
+      "epoch": 0.1952029499714357,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 3828
+    },
+    {
+      "epoch": 0.1952539434275411,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 3829
+    },
+    {
+      "epoch": 0.19530493688364647,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19535593033975188,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3831
+    },
+    {
+      "epoch": 0.19540692379585725,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8928,
+      "step": 3832
+    },
+    {
+      "epoch": 0.19545791725196265,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8735,
+      "step": 3833
+    },
+    {
+      "epoch": 0.19550891070806803,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 3834
+    },
+    {
+      "epoch": 0.19555990416417343,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 3835
+    },
+    {
+      "epoch": 0.19561089762027883,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.9008,
+      "step": 3836
+    },
+    {
+      "epoch": 0.1956618910763842,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3837
+    },
+    {
+      "epoch": 0.1957128845324896,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 3838
+    },
+    {
+      "epoch": 0.195763877988595,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 3839
+    },
+    {
+      "epoch": 0.1958148714447004,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 3840
+    },
+    {
+      "epoch": 0.19586586490080576,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 3841
+    },
+    {
+      "epoch": 0.19591685835691117,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 3842
+    },
+    {
+      "epoch": 0.19596785181301657,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 3843
+    },
+    {
+      "epoch": 0.19601884526912194,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.869,
+      "step": 3844
+    },
+    {
+      "epoch": 0.19606983872522735,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 3845
+    },
+    {
+      "epoch": 0.19612083218133272,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 3846
+    },
+    {
+      "epoch": 0.19617182563743812,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 3847
+    },
+    {
+      "epoch": 0.1962228190935435,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 3848
+    },
+    {
+      "epoch": 0.1962738125496489,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 3849
+    },
+    {
+      "epoch": 0.1963248060057543,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8867,
+      "step": 3850
+    },
+    {
+      "epoch": 0.19637579946185968,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 3851
+    },
+    {
+      "epoch": 0.19642679291796508,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8607,
+      "step": 3852
+    },
+    {
+      "epoch": 0.19647778637407046,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 3853
+    },
+    {
+      "epoch": 0.19652877983017586,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 3854
+    },
+    {
+      "epoch": 0.19657977328628123,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8772,
+      "step": 3855
+    },
+    {
+      "epoch": 0.19663076674238664,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 3856
+    },
+    {
+      "epoch": 0.19668176019849204,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 3857
+    },
+    {
+      "epoch": 0.19673275365459741,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 3858
+    },
+    {
+      "epoch": 0.19678374711070282,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 3859
+    },
+    {
+      "epoch": 0.1968347405668082,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 3860
+    },
+    {
+      "epoch": 0.1968857340229136,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 3861
+    },
+    {
+      "epoch": 0.19693672747901897,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 3862
+    },
+    {
+      "epoch": 0.19698772093512437,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 3863
+    },
+    {
+      "epoch": 0.19703871439122977,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 3864
+    },
+    {
+      "epoch": 0.19708970784733515,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8795,
+      "step": 3865
+    },
+    {
+      "epoch": 0.19714070130344055,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 3866
+    },
+    {
+      "epoch": 0.19719169475954593,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.876,
+      "step": 3867
+    },
+    {
+      "epoch": 0.19724268821565133,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 3868
+    },
+    {
+      "epoch": 0.1972936816717567,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 3869
+    },
+    {
+      "epoch": 0.1973446751278621,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8913,
+      "step": 3870
+    },
+    {
+      "epoch": 0.1973956685839675,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 3871
+    },
+    {
+      "epoch": 0.19744666204007288,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 3872
+    },
+    {
+      "epoch": 0.1974976554961783,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 3873
+    },
+    {
+      "epoch": 0.19754864895228366,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 3874
+    },
+    {
+      "epoch": 0.19759964240838906,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 3875
+    },
+    {
+      "epoch": 0.19765063586449444,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 3876
+    },
+    {
+      "epoch": 0.19770162932059984,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 3877
+    },
+    {
+      "epoch": 0.19775262277670524,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 3878
+    },
+    {
+      "epoch": 0.19780361623281062,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 3879
+    },
+    {
+      "epoch": 0.19785460968891602,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 3880
+    },
+    {
+      "epoch": 0.1979056031450214,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 3881
+    },
+    {
+      "epoch": 0.1979565966011268,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 3882
+    },
+    {
+      "epoch": 0.19800759005723217,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8731,
+      "step": 3883
+    },
+    {
+      "epoch": 0.19805858351333758,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 3884
+    },
+    {
+      "epoch": 0.19810957696944298,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 3885
+    },
+    {
+      "epoch": 0.19816057042554835,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 3886
+    },
+    {
+      "epoch": 0.19821156388165376,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 3887
+    },
+    {
+      "epoch": 0.19826255733775913,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 3888
+    },
+    {
+      "epoch": 0.19831355079386453,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 3889
+    },
+    {
+      "epoch": 0.1983645442499699,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1984155377060753,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8648,
+      "step": 3891
+    },
+    {
+      "epoch": 0.19846653116218071,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 3892
+    },
+    {
+      "epoch": 0.1985175246182861,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 3893
+    },
+    {
+      "epoch": 0.1985685180743915,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 3894
+    },
+    {
+      "epoch": 0.19861951153049687,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 3895
+    },
+    {
+      "epoch": 0.19867050498660227,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 3896
+    },
+    {
+      "epoch": 0.19872149844270764,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 3897
+    },
+    {
+      "epoch": 0.19877249189881305,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 3898
+    },
+    {
+      "epoch": 0.19882348535491845,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 3899
+    },
+    {
+      "epoch": 0.19887447881102382,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19892547226712923,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 3901
+    },
+    {
+      "epoch": 0.1989764657232346,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 3902
+    },
+    {
+      "epoch": 0.19902745917934,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 3903
+    },
+    {
+      "epoch": 0.19907845263544538,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 3904
+    },
+    {
+      "epoch": 0.19912944609155078,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 3905
+    },
+    {
+      "epoch": 0.19918043954765619,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 3906
+    },
+    {
+      "epoch": 0.19923143300376156,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 3907
+    },
+    {
+      "epoch": 0.19928242645986696,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 3908
+    },
+    {
+      "epoch": 0.19933341991597234,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8648,
+      "step": 3909
+    },
+    {
+      "epoch": 0.19938441337207774,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8978,
+      "step": 3910
+    },
+    {
+      "epoch": 0.19943540682818311,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 3911
+    },
+    {
+      "epoch": 0.19948640028428852,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 3912
+    },
+    {
+      "epoch": 0.19953739374039392,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 3913
+    },
+    {
+      "epoch": 0.1995883871964993,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 3914
+    },
+    {
+      "epoch": 0.1996393806526047,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 3915
+    },
+    {
+      "epoch": 0.19969037410871007,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 3916
+    },
+    {
+      "epoch": 0.19974136756481548,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 3917
+    },
+    {
+      "epoch": 0.19979236102092085,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 3918
+    },
+    {
+      "epoch": 0.19984335447702625,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 3919
+    },
+    {
+      "epoch": 0.19989434793313166,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19994534138923703,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 3921
+    },
+    {
+      "epoch": 0.19999633484534243,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2000473283014478,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2000983217575532,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 3924
+    },
+    {
+      "epoch": 0.20014931521365859,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8796,
+      "step": 3925
+    },
+    {
+      "epoch": 0.200200308669764,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8839,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2002513021258694,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 3927
+    },
+    {
+      "epoch": 0.20030229558197477,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 3928
+    },
+    {
+      "epoch": 0.20035328903808017,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8882,
+      "step": 3929
+    },
+    {
+      "epoch": 0.20040428249418554,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 3930
+    },
+    {
+      "epoch": 0.20045527595029095,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 3931
+    },
+    {
+      "epoch": 0.20050626940639632,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 3932
+    },
+    {
+      "epoch": 0.20055726286250172,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 3933
+    },
+    {
+      "epoch": 0.20060825631860713,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 3934
+    },
+    {
+      "epoch": 0.2006592497747125,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8798,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2007102432308179,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 3936
+    },
+    {
+      "epoch": 0.20076123668692328,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8949,
+      "step": 3937
+    },
+    {
+      "epoch": 0.20081223014302868,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8707,
+      "step": 3938
+    },
+    {
+      "epoch": 0.20086322359913406,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 3939
+    },
+    {
+      "epoch": 0.20091421705523946,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 3940
+    },
+    {
+      "epoch": 0.20096521051134486,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 3941
+    },
+    {
+      "epoch": 0.20101620396745024,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 3942
+    },
+    {
+      "epoch": 0.20106719742355564,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 3943
+    },
+    {
+      "epoch": 0.201118190879661,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 3944
+    },
+    {
+      "epoch": 0.20116918433576642,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2012201777918718,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 3946
+    },
+    {
+      "epoch": 0.2012711712479772,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2013221647040826,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 3948
+    },
+    {
+      "epoch": 0.20137315816018797,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 3949
+    },
+    {
+      "epoch": 0.20142415161629337,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8783,
+      "step": 3950
+    },
+    {
+      "epoch": 0.20147514507239875,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 3951
+    },
+    {
+      "epoch": 0.20152613852850415,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8767,
+      "step": 3952
+    },
+    {
+      "epoch": 0.20157713198460953,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 3953
+    },
+    {
+      "epoch": 0.20162812544071493,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 3954
+    },
+    {
+      "epoch": 0.20167911889682033,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2017301123529257,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 3956
+    },
+    {
+      "epoch": 0.2017811058090311,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8645,
+      "step": 3957
+    },
+    {
+      "epoch": 0.20183209926513648,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 3958
+    },
+    {
+      "epoch": 0.20188309272124189,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 3959
+    },
+    {
+      "epoch": 0.20193408617734726,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8569,
+      "step": 3960
+    },
+    {
+      "epoch": 0.20198507963345266,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8784,
+      "step": 3961
+    },
+    {
+      "epoch": 0.20203607308955807,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 3962
+    },
+    {
+      "epoch": 0.20208706654566344,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8924,
+      "step": 3963
+    },
+    {
+      "epoch": 0.20213806000176884,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 3964
+    },
+    {
+      "epoch": 0.20218905345787422,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 3965
+    },
+    {
+      "epoch": 0.20224004691397962,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 3966
+    },
+    {
+      "epoch": 0.202291040370085,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2023420338261904,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2023930272822958,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 3969
+    },
+    {
+      "epoch": 0.20244402073840118,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 3970
+    },
+    {
+      "epoch": 0.20249501419450658,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 3971
+    },
+    {
+      "epoch": 0.20254600765061195,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 3972
+    },
+    {
+      "epoch": 0.20259700110671736,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 3973
+    },
+    {
+      "epoch": 0.20264799456282273,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 3974
+    },
+    {
+      "epoch": 0.20269898801892813,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 3975
+    },
+    {
+      "epoch": 0.20274998147503354,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2028009749311389,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2028519683872443,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8599,
+      "step": 3978
+    },
+    {
+      "epoch": 0.2029029618433497,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2029539552994551,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 3980
+    },
+    {
+      "epoch": 0.20300494875556047,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 3981
+    },
+    {
+      "epoch": 0.20305594221166587,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 3982
+    },
+    {
+      "epoch": 0.20310693566777127,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.864,
+      "step": 3983
+    },
+    {
+      "epoch": 0.20315792912387665,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 3984
+    },
+    {
+      "epoch": 0.20320892257998205,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8648,
+      "step": 3985
+    },
+    {
+      "epoch": 0.20325991603608742,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 3986
+    },
+    {
+      "epoch": 0.20331090949219283,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2033619029482982,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2034128964044036,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8745,
+      "step": 3989
+    },
+    {
+      "epoch": 0.203463889860509,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8622,
+      "step": 3990
+    },
+    {
+      "epoch": 0.20351488331661438,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 3991
+    },
+    {
+      "epoch": 0.20356587677271978,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8919,
+      "step": 3992
+    },
+    {
+      "epoch": 0.20361687022882516,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 3993
+    },
+    {
+      "epoch": 0.20366786368493056,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 3994
+    },
+    {
+      "epoch": 0.20371885714103594,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 3995
+    },
+    {
+      "epoch": 0.20376985059714134,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 3996
+    },
+    {
+      "epoch": 0.20382084405324674,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 3997
+    },
+    {
+      "epoch": 0.20387183750935212,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8525,
+      "step": 3998
+    },
+    {
+      "epoch": 0.20392283096545752,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2039738244215629,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2040248178776683,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 4001
+    },
+    {
+      "epoch": 0.20407581133377367,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 4002
+    },
+    {
+      "epoch": 0.20412680478987907,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 4003
+    },
+    {
+      "epoch": 0.20417779824598448,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 4004
+    },
+    {
+      "epoch": 0.20422879170208985,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 4005
+    },
+    {
+      "epoch": 0.20427978515819525,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 4006
+    },
+    {
+      "epoch": 0.20433077861430063,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 4007
+    },
+    {
+      "epoch": 0.20438177207040603,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2044327655265114,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8727,
+      "step": 4009
+    },
+    {
+      "epoch": 0.2044837589826168,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2045347524387222,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2045857458948276,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 4012
+    },
+    {
+      "epoch": 0.204636739350933,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 4013
+    },
+    {
+      "epoch": 0.20468773280703836,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 4014
+    },
+    {
+      "epoch": 0.20473872626314377,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 4015
+    },
+    {
+      "epoch": 0.20478971971924914,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8445,
+      "step": 4016
+    },
+    {
+      "epoch": 0.20484071317535454,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 4017
+    },
+    {
+      "epoch": 0.20489170663145995,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 4018
+    },
+    {
+      "epoch": 0.20494270008756532,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 4019
+    },
+    {
+      "epoch": 0.20499369354367072,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.892,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2050446869997761,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2050956804558815,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 4022
+    },
+    {
+      "epoch": 0.20514667391198688,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 4023
+    },
+    {
+      "epoch": 0.20519766736809228,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.846,
+      "step": 4024
+    },
+    {
+      "epoch": 0.20524866082419768,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 4025
+    },
+    {
+      "epoch": 0.20529965428030306,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 4026
+    },
+    {
+      "epoch": 0.20535064773640846,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8737,
+      "step": 4027
+    },
+    {
+      "epoch": 0.20540164119251383,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 4028
+    },
+    {
+      "epoch": 0.20545263464861924,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2055036281047246,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 4030
+    },
+    {
+      "epoch": 0.20555462156083001,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8762,
+      "step": 4031
+    },
+    {
+      "epoch": 0.20560561501693542,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2056566084730408,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2057076019291462,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 4034
+    },
+    {
+      "epoch": 0.20575859538525157,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4035
+    },
+    {
+      "epoch": 0.20580958884135697,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 4036
+    },
+    {
+      "epoch": 0.20586058229746235,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 4037
+    },
+    {
+      "epoch": 0.20591157575356775,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 4038
+    },
+    {
+      "epoch": 0.20596256920967315,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 4039
+    },
+    {
+      "epoch": 0.20601356266577853,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 4040
+    },
+    {
+      "epoch": 0.20606455612188393,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8908,
+      "step": 4041
+    },
+    {
+      "epoch": 0.2061155495779893,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8742,
+      "step": 4042
+    },
+    {
+      "epoch": 0.2061665430340947,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 4043
+    },
+    {
+      "epoch": 0.20621753649020008,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8814,
+      "step": 4044
+    },
+    {
+      "epoch": 0.20626852994630548,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2063195234024109,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 4046
+    },
+    {
+      "epoch": 0.20637051685851626,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8774,
+      "step": 4047
+    },
+    {
+      "epoch": 0.20642151031462166,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8562,
+      "step": 4048
+    },
+    {
+      "epoch": 0.20647250377072704,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 4049
+    },
+    {
+      "epoch": 0.20652349722683244,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 4050
+    },
+    {
+      "epoch": 0.20657449068293782,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 4051
+    },
+    {
+      "epoch": 0.20662548413904322,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 4052
+    },
+    {
+      "epoch": 0.20667647759514862,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 4053
+    },
+    {
+      "epoch": 0.206727471051254,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2067784645073594,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4055
+    },
+    {
+      "epoch": 0.20682945796346477,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 4056
+    },
+    {
+      "epoch": 0.20688045141957018,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4057
+    },
+    {
+      "epoch": 0.20693144487567555,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 4058
+    },
+    {
+      "epoch": 0.20698243833178095,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 4059
+    },
+    {
+      "epoch": 0.20703343178788636,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 4060
+    },
+    {
+      "epoch": 0.20708442524399173,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 4061
+    },
+    {
+      "epoch": 0.20713541870009713,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2071864121562025,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.858,
+      "step": 4063
+    },
+    {
+      "epoch": 0.2072374056123079,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2072883990684133,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2073393925245187,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2073903859806241,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 4067
+    },
+    {
+      "epoch": 0.20744137943672947,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 4068
+    },
+    {
+      "epoch": 0.20749237289283487,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8913,
+      "step": 4069
+    },
+    {
+      "epoch": 0.20754336634894024,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 4070
+    },
+    {
+      "epoch": 0.20759435980504565,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 4071
+    },
+    {
+      "epoch": 0.20764535326115102,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 4072
+    },
+    {
+      "epoch": 0.20769634671725642,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 4073
+    },
+    {
+      "epoch": 0.20774734017336183,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8566,
+      "step": 4074
+    },
+    {
+      "epoch": 0.2077983336294672,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8644,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2078493270855726,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 4076
+    },
+    {
+      "epoch": 0.20790032054167798,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8486,
+      "step": 4077
+    },
+    {
+      "epoch": 0.20795131399778338,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 4078
+    },
+    {
+      "epoch": 0.20800230745388876,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8587,
+      "step": 4079
+    },
+    {
+      "epoch": 0.20805330090999416,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 4080
+    },
+    {
+      "epoch": 0.20810429436609956,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.9105,
+      "step": 4081
+    },
+    {
+      "epoch": 0.20815528782220494,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 4082
+    },
+    {
+      "epoch": 0.20820628127831034,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 4083
+    },
+    {
+      "epoch": 0.20825727473441571,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8693,
+      "step": 4084
+    },
+    {
+      "epoch": 0.20830826819052112,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2083592616466265,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2084102551027319,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2084612485588373,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 4088
+    },
+    {
+      "epoch": 0.20851224201494267,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 4089
+    },
+    {
+      "epoch": 0.20856323547104808,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 4090
+    },
+    {
+      "epoch": 0.20861422892715345,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8804,
+      "step": 4091
+    },
+    {
+      "epoch": 0.20866522238325885,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.858,
+      "step": 4092
+    },
+    {
+      "epoch": 0.20871621583936423,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 4093
+    },
+    {
+      "epoch": 0.20876720929546963,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 4094
+    },
+    {
+      "epoch": 0.20881820275157503,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2088691962076804,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2089201896637858,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4097
+    },
+    {
+      "epoch": 0.20897118311989119,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2090221765759966,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 4099
+    },
+    {
+      "epoch": 0.20907317003210196,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 4100
+    },
+    {
+      "epoch": 0.20912416348820737,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 4101
+    },
+    {
+      "epoch": 0.20917515694431277,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 4102
+    },
+    {
+      "epoch": 0.20922615040041814,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 4103
+    },
+    {
+      "epoch": 0.20927714385652355,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 4104
+    },
+    {
+      "epoch": 0.20932813731262892,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 4105
+    },
+    {
+      "epoch": 0.20937913076873432,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2094301242248397,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2094811176809451,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2095321111370505,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 4109
+    },
+    {
+      "epoch": 0.20958310459315588,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 4110
+    },
+    {
+      "epoch": 0.20963409804926128,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 4111
+    },
+    {
+      "epoch": 0.20968509150536666,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 4112
+    },
+    {
+      "epoch": 0.20973608496147206,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 4113
+    },
+    {
+      "epoch": 0.20978707841757743,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 4114
+    },
+    {
+      "epoch": 0.20983807187368284,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 4115
+    },
+    {
+      "epoch": 0.20988906532978824,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2099400587858936,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 4117
+    },
+    {
+      "epoch": 0.20999105224199902,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2100420456981044,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2100930391542098,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21014403261031517,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 4121
+    },
+    {
+      "epoch": 0.21019502606642057,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 4122
+    },
+    {
+      "epoch": 0.21024601952252597,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8731,
+      "step": 4123
+    },
+    {
+      "epoch": 0.21029701297863135,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 4124
+    },
+    {
+      "epoch": 0.21034800643473675,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8713,
+      "step": 4125
+    },
+    {
+      "epoch": 0.21039899989084213,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8781,
+      "step": 4126
+    },
+    {
+      "epoch": 0.21044999334694753,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2105009868030529,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2105519802591583,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2106029737152637,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21065396717136908,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8885,
+      "step": 4131
+    },
+    {
+      "epoch": 0.21070496062747449,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 4132
+    },
+    {
+      "epoch": 0.21075595408357986,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 4133
+    },
+    {
+      "epoch": 0.21080694753968526,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 4134
+    },
+    {
+      "epoch": 0.21085794099579064,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 4135
+    },
+    {
+      "epoch": 0.21090893445189604,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 4136
+    },
+    {
+      "epoch": 0.21095992790800144,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 4137
+    },
+    {
+      "epoch": 0.21101092136410682,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 4138
+    },
+    {
+      "epoch": 0.21106191482021222,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2111129082763176,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 4140
+    },
+    {
+      "epoch": 0.211163901732423,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 4141
+    },
+    {
+      "epoch": 0.21121489518852837,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8738,
+      "step": 4142
+    },
+    {
+      "epoch": 0.21126588864463378,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 4143
+    },
+    {
+      "epoch": 0.21131688210073918,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4144
+    },
+    {
+      "epoch": 0.21136787555684455,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4145
+    },
+    {
+      "epoch": 0.21141886901294996,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 4146
+    },
+    {
+      "epoch": 0.21146986246905533,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 4147
+    },
+    {
+      "epoch": 0.21152085592516073,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2115718493812661,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 4149
+    },
+    {
+      "epoch": 0.2116228428373715,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2116738362934769,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2117248297495823,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2117758232056877,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 4153
+    },
+    {
+      "epoch": 0.21182681666179307,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4154
+    },
+    {
+      "epoch": 0.21187781011789847,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8837,
+      "step": 4155
+    },
+    {
+      "epoch": 0.21192880357400384,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 4156
+    },
+    {
+      "epoch": 0.21197979703010925,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 4157
+    },
+    {
+      "epoch": 0.21203079048621465,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4158
+    },
+    {
+      "epoch": 0.21208178394232002,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 4159
+    },
+    {
+      "epoch": 0.21213277739842543,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2121837708545308,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 4161
+    },
+    {
+      "epoch": 0.2122347643106362,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 4162
+    },
+    {
+      "epoch": 0.21228575776674158,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 4163
+    },
+    {
+      "epoch": 0.21233675122284698,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 4164
+    },
+    {
+      "epoch": 0.21238774467895238,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 4165
+    },
+    {
+      "epoch": 0.21243873813505776,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8996,
+      "step": 4166
+    },
+    {
+      "epoch": 0.21248973159116316,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 4167
+    },
+    {
+      "epoch": 0.21254072504726854,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 4168
+    },
+    {
+      "epoch": 0.21259171850337394,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2126427119594793,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 4170
+    },
+    {
+      "epoch": 0.21269370541558472,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.878,
+      "step": 4171
+    },
+    {
+      "epoch": 0.21274469887169012,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2127956923277955,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2128466857839009,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 4174
+    },
+    {
+      "epoch": 0.21289767924000627,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 4175
+    },
+    {
+      "epoch": 0.21294867269611167,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 4176
+    },
+    {
+      "epoch": 0.21299966615221705,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 4177
+    },
+    {
+      "epoch": 0.21305065960832245,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 4178
+    },
+    {
+      "epoch": 0.21310165306442785,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 4179
+    },
+    {
+      "epoch": 0.21315264652053323,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 4180
+    },
+    {
+      "epoch": 0.21320363997663863,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 4181
+    },
+    {
+      "epoch": 0.213254633432744,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8986,
+      "step": 4182
+    },
+    {
+      "epoch": 0.2133056268888494,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 4183
+    },
+    {
+      "epoch": 0.21335662034495478,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8638,
+      "step": 4184
+    },
+    {
+      "epoch": 0.2134076138010602,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8852,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2134586072571656,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4186
+    },
+    {
+      "epoch": 0.21350960071327096,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 4187
+    },
+    {
+      "epoch": 0.21356059416937637,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 4188
+    },
+    {
+      "epoch": 0.21361158762548174,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 4189
+    },
+    {
+      "epoch": 0.21366258108158714,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21371357453769252,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 4191
+    },
+    {
+      "epoch": 0.21376456799379792,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 4192
+    },
+    {
+      "epoch": 0.21381556144990332,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 4193
+    },
+    {
+      "epoch": 0.2138665549060087,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8649,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2139175483621141,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 4195
+    },
+    {
+      "epoch": 0.21396854181821948,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 4196
+    },
+    {
+      "epoch": 0.21401953527432488,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 4197
+    },
+    {
+      "epoch": 0.21407052873043025,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 4198
+    },
+    {
+      "epoch": 0.21412152218653566,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 4199
+    },
+    {
+      "epoch": 0.21417251564264106,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8588,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21422350909874643,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 4201
+    },
+    {
+      "epoch": 0.21427450255485184,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 4202
+    },
+    {
+      "epoch": 0.2143254960109572,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 4203
+    },
+    {
+      "epoch": 0.21437648946706261,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 4204
+    },
+    {
+      "epoch": 0.214427482923168,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8727,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2144784763792734,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.877,
+      "step": 4206
+    },
+    {
+      "epoch": 0.2145294698353788,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 4207
+    },
+    {
+      "epoch": 0.21458046329148417,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 4208
+    },
+    {
+      "epoch": 0.21463145674758957,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 4209
+    },
+    {
+      "epoch": 0.21468245020369495,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8645,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21473344365980035,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 4211
+    },
+    {
+      "epoch": 0.21478443711590572,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 4212
+    },
+    {
+      "epoch": 0.21483543057201113,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 4213
+    },
+    {
+      "epoch": 0.21488642402811653,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2149374174842219,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.86,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2149884109403273,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 4216
+    },
+    {
+      "epoch": 0.21503940439643268,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 4217
+    },
+    {
+      "epoch": 0.21509039785253808,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8759,
+      "step": 4218
+    },
+    {
+      "epoch": 0.21514139130864346,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 4219
+    },
+    {
+      "epoch": 0.21519238476474886,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21524337822085426,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 4221
+    },
+    {
+      "epoch": 0.21529437167695964,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 4222
+    },
+    {
+      "epoch": 0.21534536513306504,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 4223
+    },
+    {
+      "epoch": 0.21539635858917042,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 4224
+    },
+    {
+      "epoch": 0.21544735204527582,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2154983455013812,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 4226
+    },
+    {
+      "epoch": 0.2155493389574866,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 4227
+    },
+    {
+      "epoch": 0.215600332413592,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.843,
+      "step": 4228
+    },
+    {
+      "epoch": 0.21565132586969737,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 4229
+    },
+    {
+      "epoch": 0.21570231932580278,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 4230
+    },
+    {
+      "epoch": 0.21575331278190815,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 4231
+    },
+    {
+      "epoch": 0.21580430623801355,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 4232
+    },
+    {
+      "epoch": 0.21585529969411893,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 4233
+    },
+    {
+      "epoch": 0.21590629315022433,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8894,
+      "step": 4234
+    },
+    {
+      "epoch": 0.21595728660632973,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2160082800624351,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 4236
+    },
+    {
+      "epoch": 0.2160592735185405,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2161102669746459,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2161612604307513,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 4239
+    },
+    {
+      "epoch": 0.21621225388685666,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 4240
+    },
+    {
+      "epoch": 0.21626324734296207,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 4241
+    },
+    {
+      "epoch": 0.21631424079906747,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 4242
+    },
+    {
+      "epoch": 0.21636523425517284,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8654,
+      "step": 4243
+    },
+    {
+      "epoch": 0.21641622771127825,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 4244
+    },
+    {
+      "epoch": 0.21646722116738362,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 4245
+    },
+    {
+      "epoch": 0.21651821462348902,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2165692080795944,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 4247
+    },
+    {
+      "epoch": 0.2166202015356998,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2166711949918052,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 4249
+    },
+    {
+      "epoch": 0.21672218844791058,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21677318190401598,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 4251
+    },
+    {
+      "epoch": 0.21682417536012136,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 4252
+    },
+    {
+      "epoch": 0.21687516881622676,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8676,
+      "step": 4253
+    },
+    {
+      "epoch": 0.21692616227233213,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 4254
+    },
+    {
+      "epoch": 0.21697715572843754,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8411,
+      "step": 4255
+    },
+    {
+      "epoch": 0.21702814918454294,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 4256
+    },
+    {
+      "epoch": 0.21707914264064831,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 4257
+    },
+    {
+      "epoch": 0.21713013609675372,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2171811295528591,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2172321230089645,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4260
+    },
+    {
+      "epoch": 0.21728311646506987,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 4261
+    },
+    {
+      "epoch": 0.21733410992117527,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 4262
+    },
+    {
+      "epoch": 0.21738510337728068,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 4263
+    },
+    {
+      "epoch": 0.21743609683338605,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 4264
+    },
+    {
+      "epoch": 0.21748709028949145,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 4265
+    },
+    {
+      "epoch": 0.21753808374559683,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 4266
+    },
+    {
+      "epoch": 0.21758907720170223,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2176400706578076,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.879,
+      "step": 4268
+    },
+    {
+      "epoch": 0.217691064113913,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 4269
+    },
+    {
+      "epoch": 0.2177420575700184,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21779305102612379,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2178440444822292,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 4272
+    },
+    {
+      "epoch": 0.21789503793833456,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 4273
+    },
+    {
+      "epoch": 0.21794603139443997,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 4274
+    },
+    {
+      "epoch": 0.21799702485054534,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 4275
+    },
+    {
+      "epoch": 0.21804801830665074,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 4276
+    },
+    {
+      "epoch": 0.21809901176275615,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4277
+    },
+    {
+      "epoch": 0.21815000521886152,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 4278
+    },
+    {
+      "epoch": 0.21820099867496692,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2182519921310723,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8615,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2183029855871777,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 4281
+    },
+    {
+      "epoch": 0.21835397904328308,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 4282
+    },
+    {
+      "epoch": 0.21840497249938848,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 4283
+    },
+    {
+      "epoch": 0.21845596595549388,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 4284
+    },
+    {
+      "epoch": 0.21850695941159926,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8897,
+      "step": 4285
+    },
+    {
+      "epoch": 0.21855795286770466,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 4286
+    },
+    {
+      "epoch": 0.21860894632381003,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8725,
+      "step": 4287
+    },
+    {
+      "epoch": 0.21865993977991544,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2187109332360208,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2187619266921262,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21881292014823162,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 4291
+    },
+    {
+      "epoch": 0.218863913604337,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 4292
+    },
+    {
+      "epoch": 0.2189149070604424,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 4293
+    },
+    {
+      "epoch": 0.21896590051654777,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 4294
+    },
+    {
+      "epoch": 0.21901689397265317,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 4295
+    },
+    {
+      "epoch": 0.21906788742875855,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4296
+    },
+    {
+      "epoch": 0.21911888088486395,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8609,
+      "step": 4297
+    },
+    {
+      "epoch": 0.21916987434096935,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 4298
+    },
+    {
+      "epoch": 0.21922086779707473,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8724,
+      "step": 4299
+    },
+    {
+      "epoch": 0.21927186125318013,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 4300
+    },
+    {
+      "epoch": 0.2193228547092855,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2193738481653909,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 4302
+    },
+    {
+      "epoch": 0.21942484162149628,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 4303
+    },
+    {
+      "epoch": 0.21947583507760168,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8657,
+      "step": 4304
+    },
+    {
+      "epoch": 0.21952682853370709,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 4305
+    },
+    {
+      "epoch": 0.21957782198981246,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 4306
+    },
+    {
+      "epoch": 0.21962881544591786,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 4307
+    },
+    {
+      "epoch": 0.21967980890202324,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 4308
+    },
+    {
+      "epoch": 0.21973080235812864,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 4309
+    },
+    {
+      "epoch": 0.21978179581423402,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 4310
+    },
+    {
+      "epoch": 0.21983278927033942,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.852,
+      "step": 4311
+    },
+    {
+      "epoch": 0.21988378272644482,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2199347761825502,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 4313
+    },
+    {
+      "epoch": 0.2199857696386556,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 4314
+    },
+    {
+      "epoch": 0.22003676309476097,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 4315
+    },
+    {
+      "epoch": 0.22008775655086638,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 4316
+    },
+    {
+      "epoch": 0.22013875000697175,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8566,
+      "step": 4317
+    },
+    {
+      "epoch": 0.22018974346307715,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8754,
+      "step": 4318
+    },
+    {
+      "epoch": 0.22024073691918256,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 4319
+    },
+    {
+      "epoch": 0.22029173037528793,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 4320
+    },
+    {
+      "epoch": 0.22034272383139333,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2203937172874987,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 4322
+    },
+    {
+      "epoch": 0.2204447107436041,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8724,
+      "step": 4323
+    },
+    {
+      "epoch": 0.22049570419970949,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2205466976558149,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8629,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2205976911119203,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 4326
+    },
+    {
+      "epoch": 0.22064868456802567,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8725,
+      "step": 4327
+    },
+    {
+      "epoch": 0.22069967802413107,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 4328
+    },
+    {
+      "epoch": 0.22075067148023644,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 4329
+    },
+    {
+      "epoch": 0.22080166493634185,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 4330
+    },
+    {
+      "epoch": 0.22085265839244722,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8762,
+      "step": 4331
+    },
+    {
+      "epoch": 0.22090365184855262,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 4332
+    },
+    {
+      "epoch": 0.22095464530465803,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 4333
+    },
+    {
+      "epoch": 0.2210056387607634,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2210566322168688,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 4335
+    },
+    {
+      "epoch": 0.22110762567297418,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 4336
+    },
+    {
+      "epoch": 0.22115861912907958,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4337
+    },
+    {
+      "epoch": 0.22120961258518496,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 4338
+    },
+    {
+      "epoch": 0.22126060604129036,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8647,
+      "step": 4339
+    },
+    {
+      "epoch": 0.22131159949739576,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22136259295350114,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 4341
+    },
+    {
+      "epoch": 0.22141358640960654,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2214645798657119,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 4343
+    },
+    {
+      "epoch": 0.22151557332181732,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 4344
+    },
+    {
+      "epoch": 0.2215665667779227,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2216175602340281,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 4346
+    },
+    {
+      "epoch": 0.2216685536901335,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 4347
+    },
+    {
+      "epoch": 0.22171954714623887,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 4348
+    },
+    {
+      "epoch": 0.22177054060234427,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 4349
+    },
+    {
+      "epoch": 0.22182153405844965,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22187252751455505,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 4351
+    },
+    {
+      "epoch": 0.22192352097066043,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 4352
+    },
+    {
+      "epoch": 0.22197451442676583,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8728,
+      "step": 4353
+    },
+    {
+      "epoch": 0.22202550788287123,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 4354
+    },
+    {
+      "epoch": 0.2220765013389766,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 4355
+    },
+    {
+      "epoch": 0.222127494795082,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 4356
+    },
+    {
+      "epoch": 0.22217848825118738,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2222294817072928,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 4358
+    },
+    {
+      "epoch": 0.22228047516339816,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 4359
+    },
+    {
+      "epoch": 0.22233146861950356,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22238246207560897,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 4361
+    },
+    {
+      "epoch": 0.22243345553171434,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 4362
+    },
+    {
+      "epoch": 0.22248444898781974,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.899,
+      "step": 4363
+    },
+    {
+      "epoch": 0.22253544244392512,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 4364
+    },
+    {
+      "epoch": 0.22258643590003052,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2226374293561359,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.867,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2226884228122413,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2227394162683467,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8778,
+      "step": 4368
+    },
+    {
+      "epoch": 0.22279040972445208,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4369
+    },
+    {
+      "epoch": 0.22284140318055748,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22289239663666285,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 4371
+    },
+    {
+      "epoch": 0.22294339009276826,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 4372
+    },
+    {
+      "epoch": 0.22299438354887363,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 4373
+    },
+    {
+      "epoch": 0.22304537700497903,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 4374
+    },
+    {
+      "epoch": 0.22309637046108444,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2231473639171898,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 4376
+    },
+    {
+      "epoch": 0.22319835737329521,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8645,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2232493508294006,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 4378
+    },
+    {
+      "epoch": 0.223300344285506,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8842,
+      "step": 4379
+    },
+    {
+      "epoch": 0.22335133774161137,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 4380
+    },
+    {
+      "epoch": 0.22340233119771677,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 4381
+    },
+    {
+      "epoch": 0.22345332465382217,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 4382
+    },
+    {
+      "epoch": 0.22350431810992755,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 4383
+    },
+    {
+      "epoch": 0.22355531156603295,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 4384
+    },
+    {
+      "epoch": 0.22360630502213832,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4385
+    },
+    {
+      "epoch": 0.22365729847824373,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 4386
+    },
+    {
+      "epoch": 0.2237082919343491,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 4387
+    },
+    {
+      "epoch": 0.2237592853904545,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 4388
+    },
+    {
+      "epoch": 0.2238102788465599,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 4389
+    },
+    {
+      "epoch": 0.22386127230266528,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 4390
+    },
+    {
+      "epoch": 0.22391226575877068,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 4391
+    },
+    {
+      "epoch": 0.22396325921487606,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 4392
+    },
+    {
+      "epoch": 0.22401425267098146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 4393
+    },
+    {
+      "epoch": 0.22406524612708684,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 4394
+    },
+    {
+      "epoch": 0.22411623958319224,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8643,
+      "step": 4395
+    },
+    {
+      "epoch": 0.22416723303929764,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 4396
+    },
+    {
+      "epoch": 0.22421822649540302,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 4397
+    },
+    {
+      "epoch": 0.22426921995150842,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8635,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2243202134076138,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2243712068637192,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22442220031982457,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 4401
+    },
+    {
+      "epoch": 0.22447319377592997,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 4402
+    },
+    {
+      "epoch": 0.22452418723203538,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 4403
+    },
+    {
+      "epoch": 0.22457518068814075,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4404
+    },
+    {
+      "epoch": 0.22462617414424615,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 4405
+    },
+    {
+      "epoch": 0.22467716760035153,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 4406
+    },
+    {
+      "epoch": 0.22472816105645693,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 4407
+    },
+    {
+      "epoch": 0.2247791545125623,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2248301479686677,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 4409
+    },
+    {
+      "epoch": 0.2248811414247731,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8752,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2249321348808785,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 4411
+    },
+    {
+      "epoch": 0.2249831283369839,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 4412
+    },
+    {
+      "epoch": 0.22503412179308926,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 4413
+    },
+    {
+      "epoch": 0.22508511524919467,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 4414
+    },
+    {
+      "epoch": 0.22513610870530004,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 4415
+    },
+    {
+      "epoch": 0.22518710216140544,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 4416
+    },
+    {
+      "epoch": 0.22523809561751085,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8698,
+      "step": 4417
+    },
+    {
+      "epoch": 0.22528908907361622,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 4418
+    },
+    {
+      "epoch": 0.22534008252972162,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 4419
+    },
+    {
+      "epoch": 0.225391075985827,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2254420694419324,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 4421
+    },
+    {
+      "epoch": 0.22549306289803778,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 4422
+    },
+    {
+      "epoch": 0.22554405635414318,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 4423
+    },
+    {
+      "epoch": 0.22559504981024858,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4424
+    },
+    {
+      "epoch": 0.22564604326635396,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 4425
+    },
+    {
+      "epoch": 0.22569703672245936,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8557,
+      "step": 4426
+    },
+    {
+      "epoch": 0.22574803017856473,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 4427
+    },
+    {
+      "epoch": 0.22579902363467014,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 4428
+    },
+    {
+      "epoch": 0.2258500170907755,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 4429
+    },
+    {
+      "epoch": 0.22590101054688091,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22595200400298632,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2260029974590917,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 4432
+    },
+    {
+      "epoch": 0.2260539909151971,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8479,
+      "step": 4433
+    },
+    {
+      "epoch": 0.22610498437130247,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 4434
+    },
+    {
+      "epoch": 0.22615597782740787,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 4435
+    },
+    {
+      "epoch": 0.22620697128351325,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8738,
+      "step": 4436
+    },
+    {
+      "epoch": 0.22625796473961865,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 4437
+    },
+    {
+      "epoch": 0.22630895819572405,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 4438
+    },
+    {
+      "epoch": 0.22635995165182943,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 4439
+    },
+    {
+      "epoch": 0.22641094510793483,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2264619385640402,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8709,
+      "step": 4441
+    },
+    {
+      "epoch": 0.2265129320201456,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 4442
+    },
+    {
+      "epoch": 0.22656392547625098,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 4443
+    },
+    {
+      "epoch": 0.22661491893235639,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2266659123884618,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 4445
+    },
+    {
+      "epoch": 0.22671690584456716,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 4446
+    },
+    {
+      "epoch": 0.22676789930067257,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 4447
+    },
+    {
+      "epoch": 0.22681889275677794,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 4448
+    },
+    {
+      "epoch": 0.22686988621288334,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.86,
+      "step": 4449
+    },
+    {
+      "epoch": 0.22692087966898872,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4450
+    },
+    {
+      "epoch": 0.22697187312509412,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 4451
+    },
+    {
+      "epoch": 0.22702286658119952,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8787,
+      "step": 4452
+    },
+    {
+      "epoch": 0.2270738600373049,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 4453
+    },
+    {
+      "epoch": 0.2271248534934103,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 4454
+    },
+    {
+      "epoch": 0.22717584694951568,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 4455
+    },
+    {
+      "epoch": 0.22722684040562108,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 4456
+    },
+    {
+      "epoch": 0.22727783386172645,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 4457
+    },
+    {
+      "epoch": 0.22732882731783186,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 4458
+    },
+    {
+      "epoch": 0.22737982077393726,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 4459
+    },
+    {
+      "epoch": 0.22743081423004263,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8609,
+      "step": 4460
+    },
+    {
+      "epoch": 0.22748180768614804,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 4461
+    },
+    {
+      "epoch": 0.2275328011422534,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8645,
+      "step": 4462
+    },
+    {
+      "epoch": 0.2275837945983588,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2276347880544642,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2276857815105696,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 4465
+    },
+    {
+      "epoch": 0.227736774966675,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8608,
+      "step": 4466
+    },
+    {
+      "epoch": 0.22778776842278037,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8793,
+      "step": 4467
+    },
+    {
+      "epoch": 0.22783876187888577,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 4468
+    },
+    {
+      "epoch": 0.22788975533499115,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 4469
+    },
+    {
+      "epoch": 0.22794074879109655,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22799174224720192,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 4471
+    },
+    {
+      "epoch": 0.22804273570330733,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4472
+    },
+    {
+      "epoch": 0.22809372915941273,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 4473
+    },
+    {
+      "epoch": 0.2281447226155181,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 4474
+    },
+    {
+      "epoch": 0.2281957160716235,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 4475
+    },
+    {
+      "epoch": 0.22824670952772888,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8654,
+      "step": 4476
+    },
+    {
+      "epoch": 0.22829770298383428,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 4477
+    },
+    {
+      "epoch": 0.22834869643993966,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 4478
+    },
+    {
+      "epoch": 0.22839968989604506,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 4479
+    },
+    {
+      "epoch": 0.22845068335215046,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 4480
+    },
+    {
+      "epoch": 0.22850167680825584,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 4481
+    },
+    {
+      "epoch": 0.22855267026436124,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8726,
+      "step": 4482
+    },
+    {
+      "epoch": 0.22860366372046662,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8753,
+      "step": 4483
+    },
+    {
+      "epoch": 0.22865465717657202,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 4484
+    },
+    {
+      "epoch": 0.2287056506326774,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2287566440887828,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 4486
+    },
+    {
+      "epoch": 0.2288076375448882,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 4487
+    },
+    {
+      "epoch": 0.22885863100099357,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 4488
+    },
+    {
+      "epoch": 0.22890962445709898,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 4489
+    },
+    {
+      "epoch": 0.22896061791320435,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22901161136930975,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 4491
+    },
+    {
+      "epoch": 0.22906260482541513,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 4492
+    },
+    {
+      "epoch": 0.22911359828152053,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 4493
+    },
+    {
+      "epoch": 0.22916459173762593,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 4494
+    },
+    {
+      "epoch": 0.2292155851937313,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 4495
+    },
+    {
+      "epoch": 0.2292665786498367,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 4496
+    },
+    {
+      "epoch": 0.22931757210594209,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4497
+    },
+    {
+      "epoch": 0.2293685655620475,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 4498
+    },
+    {
+      "epoch": 0.22941955901815286,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8931,
+      "step": 4499
+    },
+    {
+      "epoch": 0.22947055247425827,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22952154593036367,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 4501
+    },
+    {
+      "epoch": 0.22957253938646904,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 4502
+    },
+    {
+      "epoch": 0.22962353284257445,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8851,
+      "step": 4503
+    },
+    {
+      "epoch": 0.22967452629867982,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 4504
+    },
+    {
+      "epoch": 0.22972551975478522,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2297765132108906,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8687,
+      "step": 4506
+    },
+    {
+      "epoch": 0.229827506666996,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 4507
+    },
+    {
+      "epoch": 0.2298785001231014,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4508
+    },
+    {
+      "epoch": 0.22992949357920678,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 4509
+    },
+    {
+      "epoch": 0.22998048703531218,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23003148049141756,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 4511
+    },
+    {
+      "epoch": 0.23008247394752296,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 4512
+    },
+    {
+      "epoch": 0.23013346740362833,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 4513
+    },
+    {
+      "epoch": 0.23018446085973374,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 4514
+    },
+    {
+      "epoch": 0.23023545431583914,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 4515
+    },
+    {
+      "epoch": 0.2302864477719445,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 4516
+    },
+    {
+      "epoch": 0.23033744122804992,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2303884346841553,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 4518
+    },
+    {
+      "epoch": 0.2304394281402607,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8863,
+      "step": 4519
+    },
+    {
+      "epoch": 0.23049042159636607,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4520
+    },
+    {
+      "epoch": 0.23054141505247147,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4521
+    },
+    {
+      "epoch": 0.23059240850857687,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 4522
+    },
+    {
+      "epoch": 0.23064340196468225,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8868,
+      "step": 4523
+    },
+    {
+      "epoch": 0.23069439542078765,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 4524
+    },
+    {
+      "epoch": 0.23074538887689303,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 4525
+    },
+    {
+      "epoch": 0.23079638233299843,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 4526
+    },
+    {
+      "epoch": 0.2308473757891038,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 4527
+    },
+    {
+      "epoch": 0.2308983692452092,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2309493627013146,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 4529
+    },
+    {
+      "epoch": 0.23100035615741998,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2310513496135254,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 4531
+    },
+    {
+      "epoch": 0.23110234306963076,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 4532
+    },
+    {
+      "epoch": 0.23115333652573616,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 4533
+    },
+    {
+      "epoch": 0.23120432998184154,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 4534
+    },
+    {
+      "epoch": 0.23125532343794694,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8635,
+      "step": 4535
+    },
+    {
+      "epoch": 0.23130631689405234,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 4536
+    },
+    {
+      "epoch": 0.23135731035015772,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 4537
+    },
+    {
+      "epoch": 0.23140830380626312,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 4538
+    },
+    {
+      "epoch": 0.2314592972623685,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2315102907184739,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23156128417457927,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 4541
+    },
+    {
+      "epoch": 0.23161227763068468,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.874,
+      "step": 4542
+    },
+    {
+      "epoch": 0.23166327108679008,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 4543
+    },
+    {
+      "epoch": 0.23171426454289545,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 4544
+    },
+    {
+      "epoch": 0.23176525799900086,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 4545
+    },
+    {
+      "epoch": 0.23181625145510623,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4546
+    },
+    {
+      "epoch": 0.23186724491121163,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 4547
+    },
+    {
+      "epoch": 0.231918238367317,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 4548
+    },
+    {
+      "epoch": 0.2319692318234224,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 4549
+    },
+    {
+      "epoch": 0.23202022527952781,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2320712187356332,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2321222121917386,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 4552
+    },
+    {
+      "epoch": 0.23217320564784397,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 4553
+    },
+    {
+      "epoch": 0.23222419910394937,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 4554
+    },
+    {
+      "epoch": 0.23227519256005474,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 4555
+    },
+    {
+      "epoch": 0.23232618601616015,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 4556
+    },
+    {
+      "epoch": 0.23237717947226555,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 4557
+    },
+    {
+      "epoch": 0.23242817292837092,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 4558
+    },
+    {
+      "epoch": 0.23247916638447633,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 4559
+    },
+    {
+      "epoch": 0.2325301598405817,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8835,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2325811532966871,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 4561
+    },
+    {
+      "epoch": 0.23263214675279248,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 4562
+    },
+    {
+      "epoch": 0.23268314020889788,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 4563
+    },
+    {
+      "epoch": 0.23273413366500328,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 4564
+    },
+    {
+      "epoch": 0.23278512712110866,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 4565
+    },
+    {
+      "epoch": 0.23283612057721406,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4566
+    },
+    {
+      "epoch": 0.23288711403331944,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 4567
+    },
+    {
+      "epoch": 0.23293810748942484,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 4568
+    },
+    {
+      "epoch": 0.23298910094553021,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4569
+    },
+    {
+      "epoch": 0.23304009440163562,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8646,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23309108785774102,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2331420813138464,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 4572
+    },
+    {
+      "epoch": 0.2331930747699518,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 4573
+    },
+    {
+      "epoch": 0.23324406822605717,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8801,
+      "step": 4574
+    },
+    {
+      "epoch": 0.23329506168216257,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 4575
+    },
+    {
+      "epoch": 0.23334605513826795,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8751,
+      "step": 4576
+    },
+    {
+      "epoch": 0.23339704859437335,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 4577
+    },
+    {
+      "epoch": 0.23344804205047875,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 4578
+    },
+    {
+      "epoch": 0.23349903550658413,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8739,
+      "step": 4579
+    },
+    {
+      "epoch": 0.23355002896268953,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8877,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2336010224187949,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 4581
+    },
+    {
+      "epoch": 0.2336520158749003,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 4582
+    },
+    {
+      "epoch": 0.23370300933100568,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8717,
+      "step": 4583
+    },
+    {
+      "epoch": 0.2337540027871111,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2338049962432165,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 4585
+    },
+    {
+      "epoch": 0.23385598969932186,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8543,
+      "step": 4586
+    },
+    {
+      "epoch": 0.23390698315542727,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 4587
+    },
+    {
+      "epoch": 0.23395797661153264,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 4588
+    },
+    {
+      "epoch": 0.23400897006763804,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 4589
+    },
+    {
+      "epoch": 0.23405996352374342,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 4590
+    },
+    {
+      "epoch": 0.23411095697984882,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 4591
+    },
+    {
+      "epoch": 0.23416195043595422,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2342129438920596,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4593
+    },
+    {
+      "epoch": 0.234263937348165,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 4594
+    },
+    {
+      "epoch": 0.23431493080427038,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 4595
+    },
+    {
+      "epoch": 0.23436592426037578,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4596
+    },
+    {
+      "epoch": 0.23441691771648115,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 4597
+    },
+    {
+      "epoch": 0.23446791117258656,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 4598
+    },
+    {
+      "epoch": 0.23451890462869196,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 4599
+    },
+    {
+      "epoch": 0.23456989808479733,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23462089154090274,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8587,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2346718849970081,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 4602
+    },
+    {
+      "epoch": 0.23472287845311351,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 4603
+    },
+    {
+      "epoch": 0.2347738719092189,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 4604
+    },
+    {
+      "epoch": 0.2348248653653243,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 4605
+    },
+    {
+      "epoch": 0.2348758588214297,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8635,
+      "step": 4606
+    },
+    {
+      "epoch": 0.23492685227753507,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 4607
+    },
+    {
+      "epoch": 0.23497784573364047,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 4608
+    },
+    {
+      "epoch": 0.23502883918974585,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 4609
+    },
+    {
+      "epoch": 0.23507983264585125,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23513082610195662,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8817,
+      "step": 4611
+    },
+    {
+      "epoch": 0.23518181955806203,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 4612
+    },
+    {
+      "epoch": 0.23523281301416743,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8528,
+      "step": 4613
+    },
+    {
+      "epoch": 0.2352838064702728,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8795,
+      "step": 4614
+    },
+    {
+      "epoch": 0.2353347999263782,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 4615
+    },
+    {
+      "epoch": 0.23538579338248358,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 4616
+    },
+    {
+      "epoch": 0.23543678683858899,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8606,
+      "step": 4617
+    },
+    {
+      "epoch": 0.23548778029469436,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 4618
+    },
+    {
+      "epoch": 0.23553877375079976,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 4619
+    },
+    {
+      "epoch": 0.23558976720690517,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23564076066301054,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 4621
+    },
+    {
+      "epoch": 0.23569175411911594,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 4622
+    },
+    {
+      "epoch": 0.23574274757522132,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 4623
+    },
+    {
+      "epoch": 0.23579374103132672,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 4624
+    },
+    {
+      "epoch": 0.2358447344874321,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2358957279435375,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8756,
+      "step": 4626
+    },
+    {
+      "epoch": 0.2359467213996429,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 4627
+    },
+    {
+      "epoch": 0.23599771485574828,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 4628
+    },
+    {
+      "epoch": 0.23604870831185368,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8784,
+      "step": 4629
+    },
+    {
+      "epoch": 0.23609970176795905,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23615069522406446,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 4631
+    },
+    {
+      "epoch": 0.23620168868016983,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 4632
+    },
+    {
+      "epoch": 0.23625268213627523,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 4633
+    },
+    {
+      "epoch": 0.23630367559238064,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 4634
+    },
+    {
+      "epoch": 0.236354669048486,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2364056625045914,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2364566559606968,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 4637
+    },
+    {
+      "epoch": 0.2365076494168022,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 4638
+    },
+    {
+      "epoch": 0.23655864287290757,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8814,
+      "step": 4639
+    },
+    {
+      "epoch": 0.23660963632901297,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23666062978511837,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 4641
+    },
+    {
+      "epoch": 0.23671162324122375,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 4642
+    },
+    {
+      "epoch": 0.23676261669732915,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8433,
+      "step": 4643
+    },
+    {
+      "epoch": 0.23681361015343452,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 4644
+    },
+    {
+      "epoch": 0.23686460360953993,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2369155970656453,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 4646
+    },
+    {
+      "epoch": 0.2369665905217507,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 4647
+    },
+    {
+      "epoch": 0.2370175839778561,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 4648
+    },
+    {
+      "epoch": 0.23706857743396148,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 4649
+    },
+    {
+      "epoch": 0.23711957089006688,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 4650
+    },
+    {
+      "epoch": 0.23717056434617226,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 4651
+    },
+    {
+      "epoch": 0.23722155780227766,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 4652
+    },
+    {
+      "epoch": 0.23727255125838304,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 4653
+    },
+    {
+      "epoch": 0.23732354471448844,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 4654
+    },
+    {
+      "epoch": 0.2373745381705938,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 4655
+    },
+    {
+      "epoch": 0.23742553162669922,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 4656
+    },
+    {
+      "epoch": 0.23747652508280462,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 4657
+    },
+    {
+      "epoch": 0.23752751853891,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2375785119950154,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 4659
+    },
+    {
+      "epoch": 0.23762950545112077,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8775,
+      "step": 4660
+    },
+    {
+      "epoch": 0.23768049890722617,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 4661
+    },
+    {
+      "epoch": 0.23773149236333155,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 4662
+    },
+    {
+      "epoch": 0.23778248581943695,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 4663
+    },
+    {
+      "epoch": 0.23783347927554235,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 4664
+    },
+    {
+      "epoch": 0.23788447273164773,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 4665
+    },
+    {
+      "epoch": 0.23793546618775313,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 4666
+    },
+    {
+      "epoch": 0.2379864596438585,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 4667
+    },
+    {
+      "epoch": 0.2380374530999639,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 4668
+    },
+    {
+      "epoch": 0.23808844655606928,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8642,
+      "step": 4669
+    },
+    {
+      "epoch": 0.23813944001217469,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2381904334682801,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8834,
+      "step": 4671
+    },
+    {
+      "epoch": 0.23824142692438546,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 4672
+    },
+    {
+      "epoch": 0.23829242038049087,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 4673
+    },
+    {
+      "epoch": 0.23834341383659624,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 4674
+    },
+    {
+      "epoch": 0.23839440729270164,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 4675
+    },
+    {
+      "epoch": 0.23844540074880702,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 4676
+    },
+    {
+      "epoch": 0.23849639420491242,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 4677
+    },
+    {
+      "epoch": 0.23854738766101782,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2385983811171232,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2386493745732286,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23870036802933398,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 4681
+    },
+    {
+      "epoch": 0.23875136148543938,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 4682
+    },
+    {
+      "epoch": 0.23880235494154475,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 4683
+    },
+    {
+      "epoch": 0.23885334839765016,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 4684
+    },
+    {
+      "epoch": 0.23890434185375556,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 4685
+    },
+    {
+      "epoch": 0.23895533530986093,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 4686
+    },
+    {
+      "epoch": 0.23900632876596634,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 4687
+    },
+    {
+      "epoch": 0.2390573222220717,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8624,
+      "step": 4688
+    },
+    {
+      "epoch": 0.2391083156781771,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 4689
+    },
+    {
+      "epoch": 0.2391593091342825,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 4690
+    },
+    {
+      "epoch": 0.2392103025903879,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 4691
+    },
+    {
+      "epoch": 0.2392612960464933,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 4692
+    },
+    {
+      "epoch": 0.23931228950259867,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8853,
+      "step": 4693
+    },
+    {
+      "epoch": 0.23936328295870407,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 4694
+    },
+    {
+      "epoch": 0.23941427641480945,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.852,
+      "step": 4695
+    },
+    {
+      "epoch": 0.23946526987091485,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 4696
+    },
+    {
+      "epoch": 0.23951626332702022,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4697
+    },
+    {
+      "epoch": 0.23956725678312563,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 4698
+    },
+    {
+      "epoch": 0.23961825023923103,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 4699
+    },
+    {
+      "epoch": 0.2396692436953364,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 4700
+    },
+    {
+      "epoch": 0.2397202371514418,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 4701
+    },
+    {
+      "epoch": 0.23977123060754718,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 4702
+    },
+    {
+      "epoch": 0.23982222406365258,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 4703
+    },
+    {
+      "epoch": 0.23987321751975796,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 4704
+    },
+    {
+      "epoch": 0.23992421097586336,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8445,
+      "step": 4705
+    },
+    {
+      "epoch": 0.23997520443196876,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8896,
+      "step": 4706
+    },
+    {
+      "epoch": 0.24002619788807414,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 4707
+    },
+    {
+      "epoch": 0.24007719134417954,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 4708
+    },
+    {
+      "epoch": 0.24012818480028492,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 4709
+    },
+    {
+      "epoch": 0.24017917825639032,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2402301717124957,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 4711
+    },
+    {
+      "epoch": 0.2402811651686011,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 4712
+    },
+    {
+      "epoch": 0.2403321586247065,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 4713
+    },
+    {
+      "epoch": 0.24038315208081187,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 4714
+    },
+    {
+      "epoch": 0.24043414553691728,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4715
+    },
+    {
+      "epoch": 0.24048513899302265,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 4716
+    },
+    {
+      "epoch": 0.24053613244912805,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 4717
+    },
+    {
+      "epoch": 0.24058712590523343,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 4718
+    },
+    {
+      "epoch": 0.24063811936133883,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 4719
+    },
+    {
+      "epoch": 0.24068911281744423,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8674,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2407401062735496,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 4721
+    },
+    {
+      "epoch": 0.240791099729655,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 4722
+    },
+    {
+      "epoch": 0.2408420931857604,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 4723
+    },
+    {
+      "epoch": 0.2408930866418658,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8639,
+      "step": 4724
+    },
+    {
+      "epoch": 0.24094408009797116,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 4725
+    },
+    {
+      "epoch": 0.24099507355407657,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 4726
+    },
+    {
+      "epoch": 0.24104606701018197,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4727
+    },
+    {
+      "epoch": 0.24109706046628734,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 4728
+    },
+    {
+      "epoch": 0.24114805392239275,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8591,
+      "step": 4729
+    },
+    {
+      "epoch": 0.24119904737849812,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24125004083460352,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 4731
+    },
+    {
+      "epoch": 0.2413010342907089,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 4732
+    },
+    {
+      "epoch": 0.2413520277468143,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 4733
+    },
+    {
+      "epoch": 0.2414030212029197,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 4734
+    },
+    {
+      "epoch": 0.24145401465902508,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 4735
+    },
+    {
+      "epoch": 0.24150500811513048,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8732,
+      "step": 4736
+    },
+    {
+      "epoch": 0.24155600157123586,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 4737
+    },
+    {
+      "epoch": 0.24160699502734126,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 4738
+    },
+    {
+      "epoch": 0.24165798848344663,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 4739
+    },
+    {
+      "epoch": 0.24170898193955204,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.861,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24175997539565744,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 4741
+    },
+    {
+      "epoch": 0.24181096885176281,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 4742
+    },
+    {
+      "epoch": 0.24186196230786822,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 4743
+    },
+    {
+      "epoch": 0.2419129557639736,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 4744
+    },
+    {
+      "epoch": 0.241963949220079,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 4745
+    },
+    {
+      "epoch": 0.24201494267618437,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 4746
+    },
+    {
+      "epoch": 0.24206593613228977,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 4747
+    },
+    {
+      "epoch": 0.24211692958839517,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 4748
+    },
+    {
+      "epoch": 0.24216792304450055,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4749
+    },
+    {
+      "epoch": 0.24221891650060595,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24226990995671133,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 4751
+    },
+    {
+      "epoch": 0.24232090341281673,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 4752
+    },
+    {
+      "epoch": 0.2423718968689221,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 4753
+    },
+    {
+      "epoch": 0.2424228903250275,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 4754
+    },
+    {
+      "epoch": 0.2424738837811329,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 4755
+    },
+    {
+      "epoch": 0.24252487723723828,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 4756
+    },
+    {
+      "epoch": 0.2425758706933437,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 4757
+    },
+    {
+      "epoch": 0.24262686414944906,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 4758
+    },
+    {
+      "epoch": 0.24267785760555446,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 4759
+    },
+    {
+      "epoch": 0.24272885106165984,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24277984451776524,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 4761
+    },
+    {
+      "epoch": 0.24283083797387064,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4762
+    },
+    {
+      "epoch": 0.24288183142997602,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8588,
+      "step": 4763
+    },
+    {
+      "epoch": 0.24293282488608142,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 4764
+    },
+    {
+      "epoch": 0.2429838183421868,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 4765
+    },
+    {
+      "epoch": 0.2430348117982922,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 4766
+    },
+    {
+      "epoch": 0.24308580525439757,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 4767
+    },
+    {
+      "epoch": 0.24313679871050298,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8638,
+      "step": 4768
+    },
+    {
+      "epoch": 0.24318779216660838,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 4769
+    },
+    {
+      "epoch": 0.24323878562271375,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24328977907881916,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 4771
+    },
+    {
+      "epoch": 0.24334077253492453,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 4772
+    },
+    {
+      "epoch": 0.24339176599102993,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 4773
+    },
+    {
+      "epoch": 0.2434427594471353,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 4774
+    },
+    {
+      "epoch": 0.2434937529032407,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 4775
+    },
+    {
+      "epoch": 0.24354474635934611,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 4776
+    },
+    {
+      "epoch": 0.2435957398154515,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 4777
+    },
+    {
+      "epoch": 0.2436467332715569,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 4778
+    },
+    {
+      "epoch": 0.24369772672766227,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 4779
+    },
+    {
+      "epoch": 0.24374872018376767,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 4780
+    },
+    {
+      "epoch": 0.24379971363987304,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 4781
+    },
+    {
+      "epoch": 0.24385070709597845,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 4782
+    },
+    {
+      "epoch": 0.24390170055208385,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8779,
+      "step": 4783
+    },
+    {
+      "epoch": 0.24395269400818922,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8756,
+      "step": 4784
+    },
+    {
+      "epoch": 0.24400368746429463,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 4785
+    },
+    {
+      "epoch": 0.2440546809204,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 4786
+    },
+    {
+      "epoch": 0.2441056743765054,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 4787
+    },
+    {
+      "epoch": 0.24415666783261078,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 4788
+    },
+    {
+      "epoch": 0.24420766128871618,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 4789
+    },
+    {
+      "epoch": 0.24425865474482159,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4790
+    },
+    {
+      "epoch": 0.24430964820092696,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 4791
+    },
+    {
+      "epoch": 0.24436064165703236,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 4792
+    },
+    {
+      "epoch": 0.24441163511313774,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 4793
+    },
+    {
+      "epoch": 0.24446262856924314,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8809,
+      "step": 4794
+    },
+    {
+      "epoch": 0.24451362202534851,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 4795
+    },
+    {
+      "epoch": 0.24456461548145392,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 4796
+    },
+    {
+      "epoch": 0.24461560893755932,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2446666023936647,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 4798
+    },
+    {
+      "epoch": 0.2447175958497701,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 4799
+    },
+    {
+      "epoch": 0.24476858930587547,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24481958276198088,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8831,
+      "step": 4801
+    },
+    {
+      "epoch": 0.24487057621808625,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 4802
+    },
+    {
+      "epoch": 0.24492156967419165,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 4803
+    },
+    {
+      "epoch": 0.24497256313029706,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 4804
+    },
+    {
+      "epoch": 0.24502355658640243,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8855,
+      "step": 4805
+    },
+    {
+      "epoch": 0.24507455004250783,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 4806
+    },
+    {
+      "epoch": 0.2451255434986132,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 4807
+    },
+    {
+      "epoch": 0.2451765369547186,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 4808
+    },
+    {
+      "epoch": 0.24522753041082399,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4809
+    },
+    {
+      "epoch": 0.2452785238669294,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2453295173230348,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 4811
+    },
+    {
+      "epoch": 0.24538051077914017,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 4812
+    },
+    {
+      "epoch": 0.24543150423524557,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 4813
+    },
+    {
+      "epoch": 0.24548249769135094,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8643,
+      "step": 4814
+    },
+    {
+      "epoch": 0.24553349114745635,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 4815
+    },
+    {
+      "epoch": 0.24558448460356172,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 4816
+    },
+    {
+      "epoch": 0.24563547805966712,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 4817
+    },
+    {
+      "epoch": 0.24568647151577253,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2457374649718779,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 4819
+    },
+    {
+      "epoch": 0.2457884584279833,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4820
+    },
+    {
+      "epoch": 0.24583945188408868,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 4821
+    },
+    {
+      "epoch": 0.24589044534019408,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 4822
+    },
+    {
+      "epoch": 0.24594143879629946,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 4823
+    },
+    {
+      "epoch": 0.24599243225240486,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 4824
+    },
+    {
+      "epoch": 0.24604342570851026,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 4825
+    },
+    {
+      "epoch": 0.24609441916461564,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 4826
+    },
+    {
+      "epoch": 0.24614541262072104,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4827
+    },
+    {
+      "epoch": 0.2461964060768264,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 4828
+    },
+    {
+      "epoch": 0.24624739953293182,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 4829
+    },
+    {
+      "epoch": 0.2462983929890372,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8666,
+      "step": 4830
+    },
+    {
+      "epoch": 0.2463493864451426,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 4831
+    },
+    {
+      "epoch": 0.246400379901248,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8721,
+      "step": 4832
+    },
+    {
+      "epoch": 0.24645137335735337,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 4833
+    },
+    {
+      "epoch": 0.24650236681345877,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 4834
+    },
+    {
+      "epoch": 0.24655336026956415,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 4835
+    },
+    {
+      "epoch": 0.24660435372566955,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 4836
+    },
+    {
+      "epoch": 0.24665534718177493,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 4837
+    },
+    {
+      "epoch": 0.24670634063788033,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 4838
+    },
+    {
+      "epoch": 0.24675733409398573,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 4839
+    },
+    {
+      "epoch": 0.2468083275500911,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 4840
+    },
+    {
+      "epoch": 0.2468593210061965,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 4841
+    },
+    {
+      "epoch": 0.24691031446230188,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 4842
+    },
+    {
+      "epoch": 0.24696130791840729,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 4843
+    },
+    {
+      "epoch": 0.24701230137451266,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 4844
+    },
+    {
+      "epoch": 0.24706329483061806,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8863,
+      "step": 4845
+    },
+    {
+      "epoch": 0.24711428828672347,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 4846
+    },
+    {
+      "epoch": 0.24716528174282884,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 4847
+    },
+    {
+      "epoch": 0.24721627519893424,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 4848
+    },
+    {
+      "epoch": 0.24726726865503962,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 4849
+    },
+    {
+      "epoch": 0.24731826211114502,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2473692555672504,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2474202490233558,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2474712424794612,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4853
+    },
+    {
+      "epoch": 0.24752223593556658,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 4854
+    },
+    {
+      "epoch": 0.24757322939167198,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8788,
+      "step": 4855
+    },
+    {
+      "epoch": 0.24762422284777735,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 4856
+    },
+    {
+      "epoch": 0.24767521630388276,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 4857
+    },
+    {
+      "epoch": 0.24772620975998813,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8777,
+      "step": 4858
+    },
+    {
+      "epoch": 0.24777720321609353,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 4859
+    },
+    {
+      "epoch": 0.24782819667219894,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2478791901283043,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2479301835844097,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 4862
+    },
+    {
+      "epoch": 0.2479811770405151,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2480321704966205,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 4864
+    },
+    {
+      "epoch": 0.24808316395272587,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 4865
+    },
+    {
+      "epoch": 0.24813415740883127,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 4866
+    },
+    {
+      "epoch": 0.24818515086493667,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8616,
+      "step": 4867
+    },
+    {
+      "epoch": 0.24823614432104205,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 4868
+    },
+    {
+      "epoch": 0.24828713777714745,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 4869
+    },
+    {
+      "epoch": 0.24833813123325282,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8643,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24838912468935823,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 4871
+    },
+    {
+      "epoch": 0.2484401181454636,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 4872
+    },
+    {
+      "epoch": 0.248491111601569,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 4873
+    },
+    {
+      "epoch": 0.2485421050576744,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 4874
+    },
+    {
+      "epoch": 0.24859309851377978,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 4875
+    },
+    {
+      "epoch": 0.24864409196988518,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 4876
+    },
+    {
+      "epoch": 0.24869508542599056,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 4877
+    },
+    {
+      "epoch": 0.24874607888209596,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 4878
+    },
+    {
+      "epoch": 0.24879707233820134,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 4879
+    },
+    {
+      "epoch": 0.24884806579430674,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8543,
+      "step": 4880
+    },
+    {
+      "epoch": 0.24889905925041214,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 4881
+    },
+    {
+      "epoch": 0.24895005270651752,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 4882
+    },
+    {
+      "epoch": 0.24900104616262292,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 4883
+    },
+    {
+      "epoch": 0.2490520396187283,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8677,
+      "step": 4884
+    },
+    {
+      "epoch": 0.2491030330748337,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 4885
+    },
+    {
+      "epoch": 0.24915402653093907,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 4886
+    },
+    {
+      "epoch": 0.24920501998704447,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 4887
+    },
+    {
+      "epoch": 0.24925601344314988,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 4888
+    },
+    {
+      "epoch": 0.24930700689925525,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 4889
+    },
+    {
+      "epoch": 0.24935800035536065,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24940899381146603,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 4891
+    },
+    {
+      "epoch": 0.24945998726757143,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 4892
+    },
+    {
+      "epoch": 0.2495109807236768,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 4893
+    },
+    {
+      "epoch": 0.2495619741797822,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 4894
+    },
+    {
+      "epoch": 0.2496129676358876,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 4895
+    },
+    {
+      "epoch": 0.249663961091993,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 4896
+    },
+    {
+      "epoch": 0.2497149545480984,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 4897
+    },
+    {
+      "epoch": 0.24976594800420376,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8377,
+      "step": 4898
+    },
+    {
+      "epoch": 0.24981694146030917,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 4899
+    },
+    {
+      "epoch": 0.24986793491641454,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 4900
+    },
+    {
+      "epoch": 0.24991892837251994,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 4901
+    },
+    {
+      "epoch": 0.24996992182862535,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 4902
+    },
+    {
+      "epoch": 0.25002091528473075,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 4903
+    },
+    {
+      "epoch": 0.2500719087408361,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2501229021969415,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2501738956530469,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 4906
+    },
+    {
+      "epoch": 0.2502248891091523,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 4907
+    },
+    {
+      "epoch": 0.2502758825652577,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 4908
+    },
+    {
+      "epoch": 0.25032687602136305,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 4909
+    },
+    {
+      "epoch": 0.2503778694774685,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25042886293357386,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 4911
+    },
+    {
+      "epoch": 0.25047985638967923,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 4912
+    },
+    {
+      "epoch": 0.2505308498457846,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 4913
+    },
+    {
+      "epoch": 0.25058184330189004,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8723,
+      "step": 4914
+    },
+    {
+      "epoch": 0.2506328367579954,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 4915
+    },
+    {
+      "epoch": 0.2506838302141008,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 4916
+    },
+    {
+      "epoch": 0.2507348236702062,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 4917
+    },
+    {
+      "epoch": 0.2507858171263116,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 4918
+    },
+    {
+      "epoch": 0.25083681058241697,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 4919
+    },
+    {
+      "epoch": 0.25088780403852234,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2509387974946278,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8874,
+      "step": 4921
+    },
+    {
+      "epoch": 0.25098979095073315,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 4922
+    },
+    {
+      "epoch": 0.2510407844068385,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 4923
+    },
+    {
+      "epoch": 0.25109177786294395,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 4924
+    },
+    {
+      "epoch": 0.25114277131904933,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2511937647751547,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 4926
+    },
+    {
+      "epoch": 0.2512447582312601,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 4927
+    },
+    {
+      "epoch": 0.2512957516873655,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 4928
+    },
+    {
+      "epoch": 0.2513467451434709,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 4929
+    },
+    {
+      "epoch": 0.25139773859957626,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2514487320556817,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 4931
+    },
+    {
+      "epoch": 0.25149972551178706,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 4932
+    },
+    {
+      "epoch": 0.25155071896789244,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2516017124239978,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 4934
+    },
+    {
+      "epoch": 0.25165270588010324,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 4935
+    },
+    {
+      "epoch": 0.2517036993362086,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 4936
+    },
+    {
+      "epoch": 0.251754692792314,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 4937
+    },
+    {
+      "epoch": 0.2518056862484194,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 4938
+    },
+    {
+      "epoch": 0.2518566797045248,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 4939
+    },
+    {
+      "epoch": 0.2519076731606302,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8772,
+      "step": 4940
+    },
+    {
+      "epoch": 0.25195866661673555,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 4941
+    },
+    {
+      "epoch": 0.252009660072841,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 4942
+    },
+    {
+      "epoch": 0.25206065352894635,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 4943
+    },
+    {
+      "epoch": 0.25211164698505173,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 4944
+    },
+    {
+      "epoch": 0.25216264044115716,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 4945
+    },
+    {
+      "epoch": 0.25221363389726253,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 4946
+    },
+    {
+      "epoch": 0.2522646273533679,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 4947
+    },
+    {
+      "epoch": 0.2523156208094733,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 4948
+    },
+    {
+      "epoch": 0.2523666142655787,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 4949
+    },
+    {
+      "epoch": 0.2524176077216841,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 4950
+    },
+    {
+      "epoch": 0.25246860117778946,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 4951
+    },
+    {
+      "epoch": 0.2525195946338949,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 4952
+    },
+    {
+      "epoch": 0.25257058809000027,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 4953
+    },
+    {
+      "epoch": 0.25262158154610564,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 4954
+    },
+    {
+      "epoch": 0.252672575002211,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 4955
+    },
+    {
+      "epoch": 0.25272356845831645,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8684,
+      "step": 4956
+    },
+    {
+      "epoch": 0.2527745619144218,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 4957
+    },
+    {
+      "epoch": 0.2528255553705272,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 4958
+    },
+    {
+      "epoch": 0.25287654882663263,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 4959
+    },
+    {
+      "epoch": 0.252927542282738,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2529785357388434,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8411,
+      "step": 4961
+    },
+    {
+      "epoch": 0.25302952919494875,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 4962
+    },
+    {
+      "epoch": 0.2530805226510542,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 4963
+    },
+    {
+      "epoch": 0.25313151610715956,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 4964
+    },
+    {
+      "epoch": 0.25318250956326493,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 4965
+    },
+    {
+      "epoch": 0.25323350301937037,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 4966
+    },
+    {
+      "epoch": 0.25328449647547574,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 4967
+    },
+    {
+      "epoch": 0.2533354899315811,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 4968
+    },
+    {
+      "epoch": 0.2533864833876865,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 4969
+    },
+    {
+      "epoch": 0.2534374768437919,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 4970
+    },
+    {
+      "epoch": 0.2534884702998973,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 4971
+    },
+    {
+      "epoch": 0.25353946375600267,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 4972
+    },
+    {
+      "epoch": 0.2535904572121081,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 4973
+    },
+    {
+      "epoch": 0.2536414506682135,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 4974
+    },
+    {
+      "epoch": 0.25369244412431885,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2537434375804242,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 4976
+    },
+    {
+      "epoch": 0.25379443103652966,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 4977
+    },
+    {
+      "epoch": 0.25384542449263503,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 4978
+    },
+    {
+      "epoch": 0.2538964179487404,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 4979
+    },
+    {
+      "epoch": 0.25394741140484584,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2539984048609512,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8761,
+      "step": 4981
+    },
+    {
+      "epoch": 0.2540493983170566,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 4982
+    },
+    {
+      "epoch": 0.25410039177316196,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 4983
+    },
+    {
+      "epoch": 0.2541513852292674,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 4984
+    },
+    {
+      "epoch": 0.25420237868537277,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 4985
+    },
+    {
+      "epoch": 0.25425337214147814,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 4986
+    },
+    {
+      "epoch": 0.25430436559758357,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 4987
+    },
+    {
+      "epoch": 0.25435535905368895,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 4988
+    },
+    {
+      "epoch": 0.2544063525097943,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 4989
+    },
+    {
+      "epoch": 0.2544573459658997,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 4990
+    },
+    {
+      "epoch": 0.2545083394220051,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 4991
+    },
+    {
+      "epoch": 0.2545593328781105,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 4992
+    },
+    {
+      "epoch": 0.2546103263342159,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 4993
+    },
+    {
+      "epoch": 0.2546613197903213,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 4994
+    },
+    {
+      "epoch": 0.2547123132464267,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 4995
+    },
+    {
+      "epoch": 0.25476330670253206,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 4996
+    },
+    {
+      "epoch": 0.25481430015863743,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 4997
+    },
+    {
+      "epoch": 0.25486529361474286,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 4998
+    },
+    {
+      "epoch": 0.25491628707084824,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8665,
+      "step": 4999
+    },
+    {
+      "epoch": 0.2549672805269536,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 5000
+    },
+    {
+      "epoch": 0.25501827398305904,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 5001
+    },
+    {
+      "epoch": 0.2550692674391644,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 5002
+    },
+    {
+      "epoch": 0.2551202608952698,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 5003
+    },
+    {
+      "epoch": 0.25517125435137517,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 5004
+    },
+    {
+      "epoch": 0.2552222478074806,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5005
+    },
+    {
+      "epoch": 0.25527324126358597,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 5006
+    },
+    {
+      "epoch": 0.25532423471969135,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5007
+    },
+    {
+      "epoch": 0.2553752281757968,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 5008
+    },
+    {
+      "epoch": 0.25542622163190215,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 5009
+    },
+    {
+      "epoch": 0.2554772150880075,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.872,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2555282085441129,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 5011
+    },
+    {
+      "epoch": 0.25557920200021833,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2556301954563237,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2556811889124291,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 5014
+    },
+    {
+      "epoch": 0.2557321823685345,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2557831758246399,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 5016
+    },
+    {
+      "epoch": 0.25583416928074526,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 5017
+    },
+    {
+      "epoch": 0.25588516273685064,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 5018
+    },
+    {
+      "epoch": 0.25593615619295607,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 5019
+    },
+    {
+      "epoch": 0.25598714964906144,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2560381431051668,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 5021
+    },
+    {
+      "epoch": 0.25608913656127225,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 5022
+    },
+    {
+      "epoch": 0.2561401300173776,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 5023
+    },
+    {
+      "epoch": 0.256191123473483,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 5024
+    },
+    {
+      "epoch": 0.25624211692958837,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 5025
+    },
+    {
+      "epoch": 0.2562931103856938,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 5026
+    },
+    {
+      "epoch": 0.2563441038417992,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 5027
+    },
+    {
+      "epoch": 0.25639509729790455,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 5028
+    },
+    {
+      "epoch": 0.25644609075401,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 5029
+    },
+    {
+      "epoch": 0.25649708421011536,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8668,
+      "step": 5030
+    },
+    {
+      "epoch": 0.25654807766622073,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 5031
+    },
+    {
+      "epoch": 0.2565990711223261,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 5032
+    },
+    {
+      "epoch": 0.25665006457843154,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 5033
+    },
+    {
+      "epoch": 0.2567010580345369,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 5034
+    },
+    {
+      "epoch": 0.2567520514906423,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 5035
+    },
+    {
+      "epoch": 0.2568030449467477,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8553,
+      "step": 5036
+    },
+    {
+      "epoch": 0.2568540384028531,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 5037
+    },
+    {
+      "epoch": 0.25690503185895847,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 5038
+    },
+    {
+      "epoch": 0.25695602531506384,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 5039
+    },
+    {
+      "epoch": 0.25700701877116927,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 5040
+    },
+    {
+      "epoch": 0.25705801222727465,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 5041
+    },
+    {
+      "epoch": 0.25710900568338,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 5042
+    },
+    {
+      "epoch": 0.25715999913948545,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 5043
+    },
+    {
+      "epoch": 0.2572109925955908,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 5044
+    },
+    {
+      "epoch": 0.2572619860516962,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 5045
+    },
+    {
+      "epoch": 0.2573129795078016,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 5046
+    },
+    {
+      "epoch": 0.257363972963907,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 5047
+    },
+    {
+      "epoch": 0.2574149664200124,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 5048
+    },
+    {
+      "epoch": 0.25746595987611776,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2575169533322232,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 5050
+    },
+    {
+      "epoch": 0.25756794678832856,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8699,
+      "step": 5051
+    },
+    {
+      "epoch": 0.25761894024443394,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 5052
+    },
+    {
+      "epoch": 0.2576699337005393,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 5053
+    },
+    {
+      "epoch": 0.25772092715664474,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 5054
+    },
+    {
+      "epoch": 0.2577719206127501,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2578229140688555,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 5056
+    },
+    {
+      "epoch": 0.2578739075249609,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 5057
+    },
+    {
+      "epoch": 0.2579249009810663,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 5058
+    },
+    {
+      "epoch": 0.25797589443717167,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 5059
+    },
+    {
+      "epoch": 0.25802688789327705,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2580778813493825,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 5061
+    },
+    {
+      "epoch": 0.25812887480548785,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 5062
+    },
+    {
+      "epoch": 0.2581798682615932,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 5063
+    },
+    {
+      "epoch": 0.25823086171769866,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 5064
+    },
+    {
+      "epoch": 0.25828185517380403,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2583328486299094,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 5066
+    },
+    {
+      "epoch": 0.2583838420860148,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 5067
+    },
+    {
+      "epoch": 0.2584348355421202,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 5068
+    },
+    {
+      "epoch": 0.2584858289982256,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 5069
+    },
+    {
+      "epoch": 0.25853682245433096,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2585878159104364,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 5071
+    },
+    {
+      "epoch": 0.25863880936654177,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 5072
+    },
+    {
+      "epoch": 0.25868980282264714,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 5073
+    },
+    {
+      "epoch": 0.2587407962787525,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 5074
+    },
+    {
+      "epoch": 0.25879178973485795,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2588427831909633,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 5076
+    },
+    {
+      "epoch": 0.2588937766470687,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 5077
+    },
+    {
+      "epoch": 0.2589447701031741,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 5078
+    },
+    {
+      "epoch": 0.2589957635592795,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 5079
+    },
+    {
+      "epoch": 0.2590467570153849,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 5080
+    },
+    {
+      "epoch": 0.25909775047149025,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8703,
+      "step": 5081
+    },
+    {
+      "epoch": 0.2591487439275957,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8764,
+      "step": 5082
+    },
+    {
+      "epoch": 0.25919973738370106,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 5083
+    },
+    {
+      "epoch": 0.25925073083980643,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 5084
+    },
+    {
+      "epoch": 0.25930172429591186,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 5085
+    },
+    {
+      "epoch": 0.25935271775201724,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2594037112081226,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 5087
+    },
+    {
+      "epoch": 0.259454704664228,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 5088
+    },
+    {
+      "epoch": 0.2595056981203334,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 5089
+    },
+    {
+      "epoch": 0.2595566915764388,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 5090
+    },
+    {
+      "epoch": 0.25960768503254417,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 5091
+    },
+    {
+      "epoch": 0.2596586784886496,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 5092
+    },
+    {
+      "epoch": 0.25970967194475497,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.8587,
+      "step": 5093
+    },
+    {
+      "epoch": 0.25976066540086035,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 5094
+    },
+    {
+      "epoch": 0.2598116588569657,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 5095
+    },
+    {
+      "epoch": 0.25986265231307115,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 5096
+    },
+    {
+      "epoch": 0.2599136457691765,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 5097
+    },
+    {
+      "epoch": 0.2599646392252819,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8682,
+      "step": 5098
+    },
+    {
+      "epoch": 0.26001563268138733,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2600666261374927,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8713,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2601176195935981,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 5101
+    },
+    {
+      "epoch": 0.26016861304970346,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8755,
+      "step": 5102
+    },
+    {
+      "epoch": 0.2602196065058089,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8625,
+      "step": 5103
+    },
+    {
+      "epoch": 0.26027059996191426,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 5104
+    },
+    {
+      "epoch": 0.26032159341801964,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 5105
+    },
+    {
+      "epoch": 0.26037258687412507,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 5106
+    },
+    {
+      "epoch": 0.26042358033023044,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 5107
+    },
+    {
+      "epoch": 0.2604745737863358,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 5108
+    },
+    {
+      "epoch": 0.2605255672424412,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 5109
+    },
+    {
+      "epoch": 0.2605765606985466,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 5110
+    },
+    {
+      "epoch": 0.260627554154652,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5111
+    },
+    {
+      "epoch": 0.26067854761075737,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 5112
+    },
+    {
+      "epoch": 0.2607295410668628,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 5113
+    },
+    {
+      "epoch": 0.2607805345229682,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 5114
+    },
+    {
+      "epoch": 0.26083152797907355,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2608825214351789,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 5116
+    },
+    {
+      "epoch": 0.26093351489128436,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 5117
+    },
+    {
+      "epoch": 0.26098450834738973,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 5118
+    },
+    {
+      "epoch": 0.2610355018034951,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 5119
+    },
+    {
+      "epoch": 0.26108649525960054,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 5120
+    },
+    {
+      "epoch": 0.2611374887157059,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 5121
+    },
+    {
+      "epoch": 0.2611884821718113,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5122
+    },
+    {
+      "epoch": 0.26123947562791666,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 5123
+    },
+    {
+      "epoch": 0.2612904690840221,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 5124
+    },
+    {
+      "epoch": 0.26134146254012747,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 5125
+    },
+    {
+      "epoch": 0.26139245599623284,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5126
+    },
+    {
+      "epoch": 0.2614434494523383,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 5127
+    },
+    {
+      "epoch": 0.26149444290844365,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8629,
+      "step": 5128
+    },
+    {
+      "epoch": 0.261545436364549,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 5129
+    },
+    {
+      "epoch": 0.2615964298206544,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2616474232767598,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2616984167328652,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 5132
+    },
+    {
+      "epoch": 0.2617494101889706,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 5133
+    },
+    {
+      "epoch": 0.261800403645076,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 5134
+    },
+    {
+      "epoch": 0.2618513971011814,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 5135
+    },
+    {
+      "epoch": 0.26190239055728676,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 5136
+    },
+    {
+      "epoch": 0.26195338401339213,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 5137
+    },
+    {
+      "epoch": 0.26200437746949756,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 5138
+    },
+    {
+      "epoch": 0.26205537092560294,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5139
+    },
+    {
+      "epoch": 0.2621063643817083,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 5140
+    },
+    {
+      "epoch": 0.26215735783781374,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8528,
+      "step": 5141
+    },
+    {
+      "epoch": 0.2622083512939191,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 5142
+    },
+    {
+      "epoch": 0.2622593447500245,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 5143
+    },
+    {
+      "epoch": 0.26231033820612987,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 5144
+    },
+    {
+      "epoch": 0.2623613316622353,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 5145
+    },
+    {
+      "epoch": 0.2624123251183407,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 5146
+    },
+    {
+      "epoch": 0.26246331857444605,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 5147
+    },
+    {
+      "epoch": 0.2625143120305515,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8722,
+      "step": 5148
+    },
+    {
+      "epoch": 0.26256530548665685,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 5149
+    },
+    {
+      "epoch": 0.2626162989427622,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2626672923988676,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 5151
+    },
+    {
+      "epoch": 0.26271828585497303,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 5152
+    },
+    {
+      "epoch": 0.2627692793110784,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 5153
+    },
+    {
+      "epoch": 0.2628202727671838,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 5154
+    },
+    {
+      "epoch": 0.2628712662232892,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 5155
+    },
+    {
+      "epoch": 0.2629222596793946,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 5156
+    },
+    {
+      "epoch": 0.26297325313549996,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 5157
+    },
+    {
+      "epoch": 0.26302424659160534,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 5158
+    },
+    {
+      "epoch": 0.26307524004771077,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 5159
+    },
+    {
+      "epoch": 0.26312623350381614,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8819,
+      "step": 5160
+    },
+    {
+      "epoch": 0.2631772269599215,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 5161
+    },
+    {
+      "epoch": 0.26322822041602695,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 5162
+    },
+    {
+      "epoch": 0.2632792138721323,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 5163
+    },
+    {
+      "epoch": 0.2633302073282377,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2633812007843431,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8525,
+      "step": 5165
+    },
+    {
+      "epoch": 0.2634321942404485,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 5166
+    },
+    {
+      "epoch": 0.2634831876965539,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 5167
+    },
+    {
+      "epoch": 0.26353418115265925,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 5168
+    },
+    {
+      "epoch": 0.2635851746087647,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 5169
+    },
+    {
+      "epoch": 0.26363616806487006,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8642,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26368716152097543,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 5171
+    },
+    {
+      "epoch": 0.2637381549770808,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 5172
+    },
+    {
+      "epoch": 0.26378914843318624,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 5173
+    },
+    {
+      "epoch": 0.2638401418892916,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 5174
+    },
+    {
+      "epoch": 0.263891135345397,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 5175
+    },
+    {
+      "epoch": 0.2639421288015024,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5176
+    },
+    {
+      "epoch": 0.2639931222576078,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8924,
+      "step": 5177
+    },
+    {
+      "epoch": 0.26404411571371317,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 5178
+    },
+    {
+      "epoch": 0.26409510916981854,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8632,
+      "step": 5179
+    },
+    {
+      "epoch": 0.264146102625924,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 5180
+    },
+    {
+      "epoch": 0.26419709608202935,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 5181
+    },
+    {
+      "epoch": 0.2642480895381347,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 5182
+    },
+    {
+      "epoch": 0.26429908299424015,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 5183
+    },
+    {
+      "epoch": 0.26435007645034553,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 5184
+    },
+    {
+      "epoch": 0.2644010699064509,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2644520633625563,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8566,
+      "step": 5186
+    },
+    {
+      "epoch": 0.2645030568186617,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 5187
+    },
+    {
+      "epoch": 0.2645540502747671,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 5188
+    },
+    {
+      "epoch": 0.26460504373087246,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 5189
+    },
+    {
+      "epoch": 0.2646560371869779,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 5190
+    },
+    {
+      "epoch": 0.26470703064308326,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 5191
+    },
+    {
+      "epoch": 0.26475802409918864,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 5192
+    },
+    {
+      "epoch": 0.264809017555294,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 5193
+    },
+    {
+      "epoch": 0.26486001101139944,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 5194
+    },
+    {
+      "epoch": 0.2649110044675048,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 5195
+    },
+    {
+      "epoch": 0.2649619979236102,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5196
+    },
+    {
+      "epoch": 0.2650129913797156,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 5197
+    },
+    {
+      "epoch": 0.265063984835821,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 5198
+    },
+    {
+      "epoch": 0.2651149782919264,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 5199
+    },
+    {
+      "epoch": 0.26516597174803175,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2652169652041372,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8634,
+      "step": 5201
+    },
+    {
+      "epoch": 0.26526795866024255,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 5202
+    },
+    {
+      "epoch": 0.26531895211634793,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 5203
+    },
+    {
+      "epoch": 0.26536994557245336,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 5204
+    },
+    {
+      "epoch": 0.26542093902855873,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2654719324846641,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 5206
+    },
+    {
+      "epoch": 0.2655229259407695,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 5207
+    },
+    {
+      "epoch": 0.2655739193968749,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 5208
+    },
+    {
+      "epoch": 0.2656249128529803,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 5209
+    },
+    {
+      "epoch": 0.26567590630908566,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2657268997651911,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 5211
+    },
+    {
+      "epoch": 0.26577789322129647,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8698,
+      "step": 5212
+    },
+    {
+      "epoch": 0.26582888667740184,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 5213
+    },
+    {
+      "epoch": 0.2658798801335072,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5214
+    },
+    {
+      "epoch": 0.26593087358961265,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 5215
+    },
+    {
+      "epoch": 0.265981867045718,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 5216
+    },
+    {
+      "epoch": 0.2660328605018234,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 5217
+    },
+    {
+      "epoch": 0.26608385395792883,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 5218
+    },
+    {
+      "epoch": 0.2661348474140342,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 5219
+    },
+    {
+      "epoch": 0.2661858408701396,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 5220
+    },
+    {
+      "epoch": 0.26623683432624495,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2662878277823504,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 5222
+    },
+    {
+      "epoch": 0.26633882123845576,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 5223
+    },
+    {
+      "epoch": 0.26638981469456113,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5224
+    },
+    {
+      "epoch": 0.26644080815066656,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 5225
+    },
+    {
+      "epoch": 0.26649180160677194,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 5226
+    },
+    {
+      "epoch": 0.2665427950628773,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 5227
+    },
+    {
+      "epoch": 0.2665937885189827,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 5228
+    },
+    {
+      "epoch": 0.2666447819750881,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 5229
+    },
+    {
+      "epoch": 0.2666957754311935,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26674676888729887,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 5231
+    },
+    {
+      "epoch": 0.2667977623434043,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 5232
+    },
+    {
+      "epoch": 0.2668487557995097,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 5233
+    },
+    {
+      "epoch": 0.26689974925561505,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 5234
+    },
+    {
+      "epoch": 0.2669507427117204,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 5235
+    },
+    {
+      "epoch": 0.26700173616782585,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 5236
+    },
+    {
+      "epoch": 0.26705272962393123,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 5237
+    },
+    {
+      "epoch": 0.2671037230800366,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 5238
+    },
+    {
+      "epoch": 0.26715471653614203,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 5239
+    },
+    {
+      "epoch": 0.2672057099922474,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5240
+    },
+    {
+      "epoch": 0.2672567034483528,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8754,
+      "step": 5241
+    },
+    {
+      "epoch": 0.26730769690445816,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 5242
+    },
+    {
+      "epoch": 0.2673586903605636,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5243
+    },
+    {
+      "epoch": 0.26740968381666896,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 5244
+    },
+    {
+      "epoch": 0.26746067727277434,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 5245
+    },
+    {
+      "epoch": 0.26751167072887977,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 5246
+    },
+    {
+      "epoch": 0.26756266418498514,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 5247
+    },
+    {
+      "epoch": 0.2676136576410905,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 5248
+    },
+    {
+      "epoch": 0.2676646510971959,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 5249
+    },
+    {
+      "epoch": 0.2677156445533013,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2677666380094067,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8553,
+      "step": 5251
+    },
+    {
+      "epoch": 0.2678176314655121,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 5252
+    },
+    {
+      "epoch": 0.2678686249216175,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 5253
+    },
+    {
+      "epoch": 0.2679196183777229,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 5254
+    },
+    {
+      "epoch": 0.26797061183382825,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 5255
+    },
+    {
+      "epoch": 0.26802160528993363,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 5256
+    },
+    {
+      "epoch": 0.26807259874603906,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5257
+    },
+    {
+      "epoch": 0.26812359220214443,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.846,
+      "step": 5258
+    },
+    {
+      "epoch": 0.2681745856582498,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 5259
+    },
+    {
+      "epoch": 0.26822557911435524,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 5260
+    },
+    {
+      "epoch": 0.2682765725704606,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 5261
+    },
+    {
+      "epoch": 0.268327566026566,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5262
+    },
+    {
+      "epoch": 0.26837855948267136,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 5263
+    },
+    {
+      "epoch": 0.2684295529387768,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 5264
+    },
+    {
+      "epoch": 0.26848054639488217,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 5265
+    },
+    {
+      "epoch": 0.26853153985098754,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 5266
+    },
+    {
+      "epoch": 0.268582533307093,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 5267
+    },
+    {
+      "epoch": 0.26863352676319835,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 5268
+    },
+    {
+      "epoch": 0.2686845202193037,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8844,
+      "step": 5269
+    },
+    {
+      "epoch": 0.2687355136754091,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 5270
+    },
+    {
+      "epoch": 0.26878650713151453,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8622,
+      "step": 5271
+    },
+    {
+      "epoch": 0.2688375005876199,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 5272
+    },
+    {
+      "epoch": 0.2688884940437253,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 5273
+    },
+    {
+      "epoch": 0.2689394874998307,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 5274
+    },
+    {
+      "epoch": 0.2689904809559361,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 5275
+    },
+    {
+      "epoch": 0.26904147441204146,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 5276
+    },
+    {
+      "epoch": 0.26909246786814683,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 5277
+    },
+    {
+      "epoch": 0.26914346132425226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 5278
+    },
+    {
+      "epoch": 0.26919445478035764,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 5279
+    },
+    {
+      "epoch": 0.269245448236463,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26929644169256844,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 5281
+    },
+    {
+      "epoch": 0.2693474351486738,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 5282
+    },
+    {
+      "epoch": 0.2693984286047792,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 5283
+    },
+    {
+      "epoch": 0.26944942206088457,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 5284
+    },
+    {
+      "epoch": 0.26950041551699,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2695514089730954,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 5286
+    },
+    {
+      "epoch": 0.26960240242920075,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 5287
+    },
+    {
+      "epoch": 0.2696533958853062,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 5288
+    },
+    {
+      "epoch": 0.26970438934141155,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 5289
+    },
+    {
+      "epoch": 0.26975538279751693,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 5290
+    },
+    {
+      "epoch": 0.2698063762536223,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 5291
+    },
+    {
+      "epoch": 0.26985736970972773,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8799,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2699083631658331,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 5293
+    },
+    {
+      "epoch": 0.2699593566219385,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 5294
+    },
+    {
+      "epoch": 0.2700103500780439,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 5295
+    },
+    {
+      "epoch": 0.2700613435341493,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 5296
+    },
+    {
+      "epoch": 0.27011233699025466,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 5297
+    },
+    {
+      "epoch": 0.27016333044636004,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.858,
+      "step": 5298
+    },
+    {
+      "epoch": 0.27021432390246547,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 5299
+    },
+    {
+      "epoch": 0.27026531735857084,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2703163108146762,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 5301
+    },
+    {
+      "epoch": 0.27036730427078165,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 5302
+    },
+    {
+      "epoch": 0.270418297726887,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 5303
+    },
+    {
+      "epoch": 0.2704692911829924,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5304
+    },
+    {
+      "epoch": 0.2705202846390978,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.873,
+      "step": 5305
+    },
+    {
+      "epoch": 0.2705712780952032,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 5306
+    },
+    {
+      "epoch": 0.2706222715513086,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 5307
+    },
+    {
+      "epoch": 0.27067326500741395,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 5308
+    },
+    {
+      "epoch": 0.2707242584635194,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 5309
+    },
+    {
+      "epoch": 0.27077525191962476,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 5310
+    },
+    {
+      "epoch": 0.27082624537573013,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 5311
+    },
+    {
+      "epoch": 0.2708772388318355,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5312
+    },
+    {
+      "epoch": 0.27092823228794094,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5313
+    },
+    {
+      "epoch": 0.2709792257440463,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 5314
+    },
+    {
+      "epoch": 0.2710302192001517,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2710812126562571,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 5316
+    },
+    {
+      "epoch": 0.2711322061123625,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 5317
+    },
+    {
+      "epoch": 0.27118319956846787,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 5318
+    },
+    {
+      "epoch": 0.27123419302457324,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 5319
+    },
+    {
+      "epoch": 0.2712851864806787,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 5320
+    },
+    {
+      "epoch": 0.27133617993678405,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8671,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2713871733928894,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8676,
+      "step": 5322
+    },
+    {
+      "epoch": 0.27143816684899486,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 5323
+    },
+    {
+      "epoch": 0.27148916030510023,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8715,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2715401537612056,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5325
+    },
+    {
+      "epoch": 0.271591147217311,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 5326
+    },
+    {
+      "epoch": 0.2716421406734164,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 5327
+    },
+    {
+      "epoch": 0.2716931341295218,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 5328
+    },
+    {
+      "epoch": 0.27174412758562716,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 5329
+    },
+    {
+      "epoch": 0.2717951210417326,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 5330
+    },
+    {
+      "epoch": 0.27184611449783797,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 5331
+    },
+    {
+      "epoch": 0.27189710795394334,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 5332
+    },
+    {
+      "epoch": 0.2719481014100487,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 5333
+    },
+    {
+      "epoch": 0.27199909486615415,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 5334
+    },
+    {
+      "epoch": 0.2720500883222595,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2721010817783649,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5336
+    },
+    {
+      "epoch": 0.2721520752344703,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 5337
+    },
+    {
+      "epoch": 0.2722030686905757,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 5338
+    },
+    {
+      "epoch": 0.2722540621466811,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 5339
+    },
+    {
+      "epoch": 0.27230505560278645,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2723560490588919,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5341
+    },
+    {
+      "epoch": 0.27240704251499726,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 5342
+    },
+    {
+      "epoch": 0.27245803597110263,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 5343
+    },
+    {
+      "epoch": 0.27250902942720806,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 5344
+    },
+    {
+      "epoch": 0.27256002288331344,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 5345
+    },
+    {
+      "epoch": 0.2726110163394188,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 5346
+    },
+    {
+      "epoch": 0.2726620097955242,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 5347
+    },
+    {
+      "epoch": 0.2727130032516296,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 5348
+    },
+    {
+      "epoch": 0.272763996707735,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8629,
+      "step": 5349
+    },
+    {
+      "epoch": 0.27281499016384037,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2728659836199458,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 5351
+    },
+    {
+      "epoch": 0.27291697707605117,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 5352
+    },
+    {
+      "epoch": 0.27296797053215655,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 5353
+    },
+    {
+      "epoch": 0.2730189639882619,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 5354
+    },
+    {
+      "epoch": 0.27306995744436735,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 5355
+    },
+    {
+      "epoch": 0.2731209509004727,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 5356
+    },
+    {
+      "epoch": 0.2731719443565781,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 5357
+    },
+    {
+      "epoch": 0.27322293781268353,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2732739312687889,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 5359
+    },
+    {
+      "epoch": 0.2733249247248943,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27337591818099966,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 5361
+    },
+    {
+      "epoch": 0.2734269116371051,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 5362
+    },
+    {
+      "epoch": 0.27347790509321046,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 5363
+    },
+    {
+      "epoch": 0.27352889854931584,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 5364
+    },
+    {
+      "epoch": 0.27357989200542127,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 5365
+    },
+    {
+      "epoch": 0.27363088546152664,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 5366
+    },
+    {
+      "epoch": 0.273681878917632,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8679,
+      "step": 5367
+    },
+    {
+      "epoch": 0.2737328723737374,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 5368
+    },
+    {
+      "epoch": 0.2737838658298428,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8782,
+      "step": 5369
+    },
+    {
+      "epoch": 0.2738348592859482,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27388585274205357,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 5371
+    },
+    {
+      "epoch": 0.273936846198159,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 5372
+    },
+    {
+      "epoch": 0.2739878396542644,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 5373
+    },
+    {
+      "epoch": 0.27403883311036975,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 5374
+    },
+    {
+      "epoch": 0.2740898265664751,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8686,
+      "step": 5375
+    },
+    {
+      "epoch": 0.27414082002258056,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 5376
+    },
+    {
+      "epoch": 0.27419181347868593,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 5377
+    },
+    {
+      "epoch": 0.2742428069347913,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 5378
+    },
+    {
+      "epoch": 0.27429380039089674,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 5379
+    },
+    {
+      "epoch": 0.2743447938470021,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2743957873031075,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 5381
+    },
+    {
+      "epoch": 0.27444678075921286,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 5382
+    },
+    {
+      "epoch": 0.2744977742153183,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 5383
+    },
+    {
+      "epoch": 0.27454876767142367,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 5384
+    },
+    {
+      "epoch": 0.27459976112752904,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8553,
+      "step": 5385
+    },
+    {
+      "epoch": 0.27465075458363447,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 5386
+    },
+    {
+      "epoch": 0.27470174803973985,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 5387
+    },
+    {
+      "epoch": 0.2747527414958452,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 5388
+    },
+    {
+      "epoch": 0.2748037349519506,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 5389
+    },
+    {
+      "epoch": 0.274854728408056,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 5390
+    },
+    {
+      "epoch": 0.2749057218641614,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 5391
+    },
+    {
+      "epoch": 0.2749567153202668,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5392
+    },
+    {
+      "epoch": 0.2750077087763722,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 5393
+    },
+    {
+      "epoch": 0.2750587022324776,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 5394
+    },
+    {
+      "epoch": 0.27510969568858296,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 5395
+    },
+    {
+      "epoch": 0.27516068914468833,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 5396
+    },
+    {
+      "epoch": 0.27521168260079376,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 5397
+    },
+    {
+      "epoch": 0.27526267605689914,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 5398
+    },
+    {
+      "epoch": 0.2753136695130045,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8879,
+      "step": 5399
+    },
+    {
+      "epoch": 0.27536466296910994,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2754156564252153,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 5401
+    },
+    {
+      "epoch": 0.2754666498813207,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 5402
+    },
+    {
+      "epoch": 0.27551764333742607,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 5403
+    },
+    {
+      "epoch": 0.2755686367935315,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 5404
+    },
+    {
+      "epoch": 0.27561963024963687,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 5405
+    },
+    {
+      "epoch": 0.27567062370574225,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 5406
+    },
+    {
+      "epoch": 0.2757216171618477,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 5407
+    },
+    {
+      "epoch": 0.27577261061795305,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 5408
+    },
+    {
+      "epoch": 0.2758236040740584,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 5409
+    },
+    {
+      "epoch": 0.2758745975301638,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5410
+    },
+    {
+      "epoch": 0.27592559098626923,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 5411
+    },
+    {
+      "epoch": 0.2759765844423746,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 5412
+    },
+    {
+      "epoch": 0.27602757789848,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 5413
+    },
+    {
+      "epoch": 0.2760785713545854,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 5414
+    },
+    {
+      "epoch": 0.2761295648106908,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 5415
+    },
+    {
+      "epoch": 0.27618055826679616,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 5416
+    },
+    {
+      "epoch": 0.27623155172290154,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 5417
+    },
+    {
+      "epoch": 0.27628254517900697,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 5418
+    },
+    {
+      "epoch": 0.27633353863511234,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 5419
+    },
+    {
+      "epoch": 0.2763845320912177,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 5420
+    },
+    {
+      "epoch": 0.27643552554732315,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 5421
+    },
+    {
+      "epoch": 0.2764865190034285,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5422
+    },
+    {
+      "epoch": 0.2765375124595339,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 5423
+    },
+    {
+      "epoch": 0.27658850591563927,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5424
+    },
+    {
+      "epoch": 0.2766394993717447,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2766904928278501,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 5426
+    },
+    {
+      "epoch": 0.27674148628395545,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 5427
+    },
+    {
+      "epoch": 0.2767924797400609,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 5428
+    },
+    {
+      "epoch": 0.27684347319616626,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8817,
+      "step": 5429
+    },
+    {
+      "epoch": 0.27689446665227163,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 5430
+    },
+    {
+      "epoch": 0.276945460108377,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 5431
+    },
+    {
+      "epoch": 0.27699645356448244,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2770474470205878,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 5433
+    },
+    {
+      "epoch": 0.2770984404766932,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 5434
+    },
+    {
+      "epoch": 0.2771494339327986,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 5435
+    },
+    {
+      "epoch": 0.277200427388904,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 5436
+    },
+    {
+      "epoch": 0.27725142084500937,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 5437
+    },
+    {
+      "epoch": 0.27730241430111474,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 5438
+    },
+    {
+      "epoch": 0.27735340775722017,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 5439
+    },
+    {
+      "epoch": 0.27740440121332555,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2774553946694309,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 5441
+    },
+    {
+      "epoch": 0.27750638812553635,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 5442
+    },
+    {
+      "epoch": 0.2775573815816417,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 5443
+    },
+    {
+      "epoch": 0.2776083750377471,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 5444
+    },
+    {
+      "epoch": 0.2776593684938525,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2777103619499579,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 5446
+    },
+    {
+      "epoch": 0.2777613554060633,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 5447
+    },
+    {
+      "epoch": 0.27781234886216866,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 5448
+    },
+    {
+      "epoch": 0.2778633423182741,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 5449
+    },
+    {
+      "epoch": 0.27791433577437946,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 5450
+    },
+    {
+      "epoch": 0.27796532923048484,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 5451
+    },
+    {
+      "epoch": 0.2780163226865902,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5452
+    },
+    {
+      "epoch": 0.27806731614269564,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 5453
+    },
+    {
+      "epoch": 0.278118309598801,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8433,
+      "step": 5454
+    },
+    {
+      "epoch": 0.2781693030549064,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.866,
+      "step": 5455
+    },
+    {
+      "epoch": 0.2782202965110118,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5456
+    },
+    {
+      "epoch": 0.2782712899671172,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 5457
+    },
+    {
+      "epoch": 0.27832228342322257,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 5458
+    },
+    {
+      "epoch": 0.27837327687932795,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 5459
+    },
+    {
+      "epoch": 0.2784242703354334,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 5460
+    },
+    {
+      "epoch": 0.27847526379153875,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8606,
+      "step": 5461
+    },
+    {
+      "epoch": 0.2785262572476441,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 5462
+    },
+    {
+      "epoch": 0.27857725070374956,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 5463
+    },
+    {
+      "epoch": 0.27862824415985493,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 5464
+    },
+    {
+      "epoch": 0.2786792376159603,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 5465
+    },
+    {
+      "epoch": 0.2787302310720657,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 5466
+    },
+    {
+      "epoch": 0.2787812245281711,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 5467
+    },
+    {
+      "epoch": 0.2788322179842765,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8558,
+      "step": 5468
+    },
+    {
+      "epoch": 0.27888321144038186,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 5469
+    },
+    {
+      "epoch": 0.2789342048964873,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5470
+    },
+    {
+      "epoch": 0.27898519835259267,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 5471
+    },
+    {
+      "epoch": 0.27903619180869804,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 5472
+    },
+    {
+      "epoch": 0.2790871852648034,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5473
+    },
+    {
+      "epoch": 0.27913817872090885,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 5474
+    },
+    {
+      "epoch": 0.2791891721770142,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 5475
+    },
+    {
+      "epoch": 0.2792401656331196,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 5476
+    },
+    {
+      "epoch": 0.279291159089225,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 5477
+    },
+    {
+      "epoch": 0.2793421525453304,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 5478
+    },
+    {
+      "epoch": 0.2793931460014358,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 5479
+    },
+    {
+      "epoch": 0.27944413945754115,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2794951329136466,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 5481
+    },
+    {
+      "epoch": 0.27954612636975196,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 5482
+    },
+    {
+      "epoch": 0.27959711982585733,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5483
+    },
+    {
+      "epoch": 0.27964811328196276,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 5484
+    },
+    {
+      "epoch": 0.27969910673806814,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 5485
+    },
+    {
+      "epoch": 0.2797501001941735,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8653,
+      "step": 5486
+    },
+    {
+      "epoch": 0.2798010936502789,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8778,
+      "step": 5487
+    },
+    {
+      "epoch": 0.2798520871063843,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 5488
+    },
+    {
+      "epoch": 0.2799030805624897,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 5489
+    },
+    {
+      "epoch": 0.27995407401859507,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2800050674747005,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 5491
+    },
+    {
+      "epoch": 0.2800560609308059,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 5492
+    },
+    {
+      "epoch": 0.28010705438691125,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8704,
+      "step": 5493
+    },
+    {
+      "epoch": 0.2801580478430166,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 5494
+    },
+    {
+      "epoch": 0.28020904129912205,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 5495
+    },
+    {
+      "epoch": 0.2802600347552274,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 5496
+    },
+    {
+      "epoch": 0.2803110282113328,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8765,
+      "step": 5497
+    },
+    {
+      "epoch": 0.28036202166743823,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 5498
+    },
+    {
+      "epoch": 0.2804130151235436,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 5499
+    },
+    {
+      "epoch": 0.280464008579649,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8712,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28051500203575436,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 5501
+    },
+    {
+      "epoch": 0.2805659954918598,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8769,
+      "step": 5502
+    },
+    {
+      "epoch": 0.28061698894796516,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 5503
+    },
+    {
+      "epoch": 0.28066798240407054,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5504
+    },
+    {
+      "epoch": 0.28071897586017597,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 5505
+    },
+    {
+      "epoch": 0.28076996931628134,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 5506
+    },
+    {
+      "epoch": 0.2808209627723867,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 5507
+    },
+    {
+      "epoch": 0.2808719562284921,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 5508
+    },
+    {
+      "epoch": 0.2809229496845975,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 5509
+    },
+    {
+      "epoch": 0.2809739431407029,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2810249365968083,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 5511
+    },
+    {
+      "epoch": 0.2810759300529137,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 5512
+    },
+    {
+      "epoch": 0.2811269235090191,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 5513
+    },
+    {
+      "epoch": 0.28117791696512445,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 5514
+    },
+    {
+      "epoch": 0.2812289104212298,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 5515
+    },
+    {
+      "epoch": 0.28127990387733526,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 5516
+    },
+    {
+      "epoch": 0.28133089733344063,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 5517
+    },
+    {
+      "epoch": 0.281381890789546,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 5518
+    },
+    {
+      "epoch": 0.28143288424565144,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 5519
+    },
+    {
+      "epoch": 0.2814838777017568,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2815348711578622,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8606,
+      "step": 5521
+    },
+    {
+      "epoch": 0.28158586461396756,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 5522
+    },
+    {
+      "epoch": 0.281636858070073,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 5523
+    },
+    {
+      "epoch": 0.28168785152617837,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 5524
+    },
+    {
+      "epoch": 0.28173884498228374,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 5525
+    },
+    {
+      "epoch": 0.2817898384383892,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 5526
+    },
+    {
+      "epoch": 0.28184083189449455,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 5527
+    },
+    {
+      "epoch": 0.2818918253505999,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5528
+    },
+    {
+      "epoch": 0.2819428188067053,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 5529
+    },
+    {
+      "epoch": 0.28199381226281073,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2820448057189161,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 5531
+    },
+    {
+      "epoch": 0.2820957991750215,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 5532
+    },
+    {
+      "epoch": 0.2821467926311269,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 5533
+    },
+    {
+      "epoch": 0.2821977860872323,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8592,
+      "step": 5534
+    },
+    {
+      "epoch": 0.28224877954333766,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 5535
+    },
+    {
+      "epoch": 0.28229977299944303,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 5536
+    },
+    {
+      "epoch": 0.28235076645554846,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 5537
+    },
+    {
+      "epoch": 0.28240175991165384,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 5538
+    },
+    {
+      "epoch": 0.2824527533677592,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 5539
+    },
+    {
+      "epoch": 0.28250374682386464,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 5540
+    },
+    {
+      "epoch": 0.28255474027997,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 5541
+    },
+    {
+      "epoch": 0.2826057337360754,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 5542
+    },
+    {
+      "epoch": 0.28265672719218077,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 5543
+    },
+    {
+      "epoch": 0.2827077206482862,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 5544
+    },
+    {
+      "epoch": 0.2827587141043916,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 5545
+    },
+    {
+      "epoch": 0.28280970756049695,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 5546
+    },
+    {
+      "epoch": 0.2828607010166024,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8823,
+      "step": 5547
+    },
+    {
+      "epoch": 0.28291169447270775,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 5548
+    },
+    {
+      "epoch": 0.28296268792881313,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5549
+    },
+    {
+      "epoch": 0.2830136813849185,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28306467484102393,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 5551
+    },
+    {
+      "epoch": 0.2831156682971293,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 5552
+    },
+    {
+      "epoch": 0.2831666617532347,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 5553
+    },
+    {
+      "epoch": 0.2832176552093401,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5554
+    },
+    {
+      "epoch": 0.2832686486654455,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8687,
+      "step": 5555
+    },
+    {
+      "epoch": 0.28331964212155086,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 5556
+    },
+    {
+      "epoch": 0.28337063557765624,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8642,
+      "step": 5557
+    },
+    {
+      "epoch": 0.28342162903376167,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.843,
+      "step": 5558
+    },
+    {
+      "epoch": 0.28347262248986704,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 5559
+    },
+    {
+      "epoch": 0.2835236159459724,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28357460940207785,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 5561
+    },
+    {
+      "epoch": 0.2836256028581832,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8905,
+      "step": 5562
+    },
+    {
+      "epoch": 0.2836765963142886,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 5563
+    },
+    {
+      "epoch": 0.283727589770394,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 5564
+    },
+    {
+      "epoch": 0.2837785832264994,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 5565
+    },
+    {
+      "epoch": 0.2838295766826048,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 5566
+    },
+    {
+      "epoch": 0.28388057013871015,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 5567
+    },
+    {
+      "epoch": 0.2839315635948156,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 5568
+    },
+    {
+      "epoch": 0.28398255705092096,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 5569
+    },
+    {
+      "epoch": 0.28403355050702633,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8609,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2840845439631317,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 5571
+    },
+    {
+      "epoch": 0.28413553741923714,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 5572
+    },
+    {
+      "epoch": 0.2841865308753425,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8607,
+      "step": 5573
+    },
+    {
+      "epoch": 0.2842375243314479,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8565,
+      "step": 5574
+    },
+    {
+      "epoch": 0.2842885177875533,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8585,
+      "step": 5575
+    },
+    {
+      "epoch": 0.2843395112436587,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 5576
+    },
+    {
+      "epoch": 0.28439050469976407,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 5577
+    },
+    {
+      "epoch": 0.28444149815586944,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8602,
+      "step": 5578
+    },
+    {
+      "epoch": 0.2844924916119749,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8736,
+      "step": 5579
+    },
+    {
+      "epoch": 0.28454348506808025,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2845944785241856,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 5581
+    },
+    {
+      "epoch": 0.28464547198029105,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 5582
+    },
+    {
+      "epoch": 0.28469646543639643,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 5583
+    },
+    {
+      "epoch": 0.2847474588925018,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 5584
+    },
+    {
+      "epoch": 0.2847984523486072,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 5585
+    },
+    {
+      "epoch": 0.2848494458047126,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 5586
+    },
+    {
+      "epoch": 0.284900439260818,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 5587
+    },
+    {
+      "epoch": 0.28495143271692336,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 5588
+    },
+    {
+      "epoch": 0.2850024261730288,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 5589
+    },
+    {
+      "epoch": 0.28505341962913416,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 5590
+    },
+    {
+      "epoch": 0.28510441308523954,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 5591
+    },
+    {
+      "epoch": 0.2851554065413449,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 5592
+    },
+    {
+      "epoch": 0.28520639999745034,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 5593
+    },
+    {
+      "epoch": 0.2852573934535557,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 5594
+    },
+    {
+      "epoch": 0.2853083869096611,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 5595
+    },
+    {
+      "epoch": 0.2853593803657665,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 5596
+    },
+    {
+      "epoch": 0.2854103738218719,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8557,
+      "step": 5597
+    },
+    {
+      "epoch": 0.2854613672779773,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 5598
+    },
+    {
+      "epoch": 0.28551236073408265,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8619,
+      "step": 5599
+    },
+    {
+      "epoch": 0.2855633541901881,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 5600
+    },
+    {
+      "epoch": 0.28561434764629345,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 5601
+    },
+    {
+      "epoch": 0.28566534110239883,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 5602
+    },
+    {
+      "epoch": 0.28571633455850426,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 5603
+    },
+    {
+      "epoch": 0.28576732801460963,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 5604
+    },
+    {
+      "epoch": 0.285818321470715,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2858693149268204,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 5606
+    },
+    {
+      "epoch": 0.2859203083829258,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 5607
+    },
+    {
+      "epoch": 0.2859713018390312,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 5608
+    },
+    {
+      "epoch": 0.28602229529513656,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 5609
+    },
+    {
+      "epoch": 0.286073288751242,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 5610
+    },
+    {
+      "epoch": 0.28612428220734737,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 5611
+    },
+    {
+      "epoch": 0.28617527566345274,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5612
+    },
+    {
+      "epoch": 0.2862262691195581,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 5613
+    },
+    {
+      "epoch": 0.28627726257566355,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 5614
+    },
+    {
+      "epoch": 0.2863282560317689,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5615
+    },
+    {
+      "epoch": 0.2863792494878743,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5616
+    },
+    {
+      "epoch": 0.28643024294397973,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 5617
+    },
+    {
+      "epoch": 0.2864812364000851,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 5618
+    },
+    {
+      "epoch": 0.2865322298561905,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8744,
+      "step": 5619
+    },
+    {
+      "epoch": 0.28658322331229585,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 5620
+    },
+    {
+      "epoch": 0.2866342167684013,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 5621
+    },
+    {
+      "epoch": 0.28668521022450666,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8563,
+      "step": 5622
+    },
+    {
+      "epoch": 0.28673620368061203,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5623
+    },
+    {
+      "epoch": 0.28678719713671746,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 5624
+    },
+    {
+      "epoch": 0.28683819059282284,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 5625
+    },
+    {
+      "epoch": 0.2868891840489282,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 5626
+    },
+    {
+      "epoch": 0.2869401775050336,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 5627
+    },
+    {
+      "epoch": 0.286991170961139,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 5628
+    },
+    {
+      "epoch": 0.2870421644172444,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 5629
+    },
+    {
+      "epoch": 0.28709315787334977,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2871441513294552,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 5631
+    },
+    {
+      "epoch": 0.2871951447855606,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 5632
+    },
+    {
+      "epoch": 0.28724613824166595,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 5633
+    },
+    {
+      "epoch": 0.2872971316977713,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 5634
+    },
+    {
+      "epoch": 0.28734812515387675,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 5635
+    },
+    {
+      "epoch": 0.28739911860998213,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 5636
+    },
+    {
+      "epoch": 0.2874501120660875,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 5637
+    },
+    {
+      "epoch": 0.28750110552219293,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 5638
+    },
+    {
+      "epoch": 0.2875520989782983,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 5639
+    },
+    {
+      "epoch": 0.2876030924344037,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 5640
+    },
+    {
+      "epoch": 0.28765408589050906,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 5641
+    },
+    {
+      "epoch": 0.2877050793466145,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8583,
+      "step": 5642
+    },
+    {
+      "epoch": 0.28775607280271986,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 5643
+    },
+    {
+      "epoch": 0.28780706625882524,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 5644
+    },
+    {
+      "epoch": 0.28785805971493067,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8613,
+      "step": 5645
+    },
+    {
+      "epoch": 0.28790905317103604,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 5646
+    },
+    {
+      "epoch": 0.2879600466271414,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 5647
+    },
+    {
+      "epoch": 0.2880110400832468,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 5648
+    },
+    {
+      "epoch": 0.2880620335393522,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 5649
+    },
+    {
+      "epoch": 0.2881130269954576,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 5650
+    },
+    {
+      "epoch": 0.288164020451563,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 5651
+    },
+    {
+      "epoch": 0.2882150139076684,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 5652
+    },
+    {
+      "epoch": 0.2882660073637738,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5653
+    },
+    {
+      "epoch": 0.28831700081987915,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5654
+    },
+    {
+      "epoch": 0.28836799427598453,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 5655
+    },
+    {
+      "epoch": 0.28841898773208996,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 5656
+    },
+    {
+      "epoch": 0.28846998118819533,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 5657
+    },
+    {
+      "epoch": 0.2885209746443007,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 5658
+    },
+    {
+      "epoch": 0.28857196810040614,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2886229615565115,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2886739550126169,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 5661
+    },
+    {
+      "epoch": 0.28872494846872226,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 5662
+    },
+    {
+      "epoch": 0.2887759419248277,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 5663
+    },
+    {
+      "epoch": 0.28882693538093307,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 5664
+    },
+    {
+      "epoch": 0.28887792883703844,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 5665
+    },
+    {
+      "epoch": 0.2889289222931439,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 5666
+    },
+    {
+      "epoch": 0.28897991574924925,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 5667
+    },
+    {
+      "epoch": 0.2890309092053546,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 5668
+    },
+    {
+      "epoch": 0.28908190266146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 5669
+    },
+    {
+      "epoch": 0.28913289611756543,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 5670
+    },
+    {
+      "epoch": 0.2891838895736708,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 2.8569,
+      "step": 5671
+    },
+    {
+      "epoch": 0.2892348830297762,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 5672
+    },
+    {
+      "epoch": 0.2892858764858816,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.843,
+      "step": 5673
+    },
+    {
+      "epoch": 0.289336869941987,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 5674
+    },
+    {
+      "epoch": 0.28938786339809236,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 5675
+    },
+    {
+      "epoch": 0.28943885685419773,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 5676
+    },
+    {
+      "epoch": 0.28948985031030317,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 5677
+    },
+    {
+      "epoch": 0.28954084376640854,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8651,
+      "step": 5678
+    },
+    {
+      "epoch": 0.2895918372225139,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 5679
+    },
+    {
+      "epoch": 0.28964283067861935,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8807,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2896938241347247,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 5681
+    },
+    {
+      "epoch": 0.2897448175908301,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 5682
+    },
+    {
+      "epoch": 0.28979581104693547,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 5683
+    },
+    {
+      "epoch": 0.2898468045030409,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 5684
+    },
+    {
+      "epoch": 0.2898977979591463,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5685
+    },
+    {
+      "epoch": 0.28994879141525165,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 5686
+    },
+    {
+      "epoch": 0.2899997848713571,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 5687
+    },
+    {
+      "epoch": 0.29005077832746246,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8562,
+      "step": 5688
+    },
+    {
+      "epoch": 0.29010177178356783,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 5689
+    },
+    {
+      "epoch": 0.2901527652396732,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29020375869577864,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5691
+    },
+    {
+      "epoch": 0.290254752151884,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 5692
+    },
+    {
+      "epoch": 0.2903057456079894,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 5693
+    },
+    {
+      "epoch": 0.2903567390640948,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 5694
+    },
+    {
+      "epoch": 0.2904077325202002,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 5695
+    },
+    {
+      "epoch": 0.29045872597630557,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 5696
+    },
+    {
+      "epoch": 0.29050971943241094,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 5697
+    },
+    {
+      "epoch": 0.29056071288851637,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8627,
+      "step": 5698
+    },
+    {
+      "epoch": 0.29061170634462175,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 5699
+    },
+    {
+      "epoch": 0.2906626998007271,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29071369325683255,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 5701
+    },
+    {
+      "epoch": 0.2907646867129379,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 5702
+    },
+    {
+      "epoch": 0.2908156801690433,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 5703
+    },
+    {
+      "epoch": 0.2908666736251487,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8741,
+      "step": 5704
+    },
+    {
+      "epoch": 0.2909176670812541,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 5705
+    },
+    {
+      "epoch": 0.2909686605373595,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 5706
+    },
+    {
+      "epoch": 0.29101965399346486,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 5707
+    },
+    {
+      "epoch": 0.2910706474495703,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 5708
+    },
+    {
+      "epoch": 0.29112164090567566,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 5709
+    },
+    {
+      "epoch": 0.29117263436178104,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2912236278178864,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 5711
+    },
+    {
+      "epoch": 0.29127462127399184,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 5712
+    },
+    {
+      "epoch": 0.2913256147300972,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 5713
+    },
+    {
+      "epoch": 0.2913766081862026,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 5714
+    },
+    {
+      "epoch": 0.291427601642308,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 5715
+    },
+    {
+      "epoch": 0.2914785950984134,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 5716
+    },
+    {
+      "epoch": 0.29152958855451877,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 5717
+    },
+    {
+      "epoch": 0.29158058201062415,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 5718
+    },
+    {
+      "epoch": 0.2916315754667296,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 5719
+    },
+    {
+      "epoch": 0.29168256892283495,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2917335623789403,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 5721
+    },
+    {
+      "epoch": 0.29178455583504576,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 5722
+    },
+    {
+      "epoch": 0.29183554929115113,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8637,
+      "step": 5723
+    },
+    {
+      "epoch": 0.2918865427472565,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 5724
+    },
+    {
+      "epoch": 0.2919375362033619,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 5725
+    },
+    {
+      "epoch": 0.2919885296594673,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 5726
+    },
+    {
+      "epoch": 0.2920395231155727,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 5727
+    },
+    {
+      "epoch": 0.29209051657167806,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 5728
+    },
+    {
+      "epoch": 0.2921415100277835,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 5729
+    },
+    {
+      "epoch": 0.29219250348388887,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 5730
+    },
+    {
+      "epoch": 0.29224349693999424,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 5731
+    },
+    {
+      "epoch": 0.2922944903960996,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 5732
+    },
+    {
+      "epoch": 0.29234548385220505,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 5733
+    },
+    {
+      "epoch": 0.2923964773083104,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 5734
+    },
+    {
+      "epoch": 0.2924474707644158,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 5735
+    },
+    {
+      "epoch": 0.2924984642205212,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 5736
+    },
+    {
+      "epoch": 0.2925494576766266,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 5737
+    },
+    {
+      "epoch": 0.292600451132732,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 5738
+    },
+    {
+      "epoch": 0.29265144458883735,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 5739
+    },
+    {
+      "epoch": 0.2927024380449428,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29275343150104816,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 5741
+    },
+    {
+      "epoch": 0.29280442495715353,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 5742
+    },
+    {
+      "epoch": 0.29285541841325896,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.871,
+      "step": 5743
+    },
+    {
+      "epoch": 0.29290641186936434,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 5744
+    },
+    {
+      "epoch": 0.2929574053254697,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8714,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2930083987815751,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 5746
+    },
+    {
+      "epoch": 0.2930593922376805,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 5747
+    },
+    {
+      "epoch": 0.2931103856937859,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 5748
+    },
+    {
+      "epoch": 0.29316137914989127,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 5749
+    },
+    {
+      "epoch": 0.2932123726059967,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29326336606210207,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8506,
+      "step": 5751
+    },
+    {
+      "epoch": 0.29331435951820745,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 5752
+    },
+    {
+      "epoch": 0.2933653529743128,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 5753
+    },
+    {
+      "epoch": 0.29341634643041825,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 5754
+    },
+    {
+      "epoch": 0.2934673398865236,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 5755
+    },
+    {
+      "epoch": 0.293518333342629,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 5756
+    },
+    {
+      "epoch": 0.29356932679873443,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 5757
+    },
+    {
+      "epoch": 0.2936203202548398,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 5758
+    },
+    {
+      "epoch": 0.2936713137109452,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 5759
+    },
+    {
+      "epoch": 0.29372230716705056,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 5760
+    },
+    {
+      "epoch": 0.293773300623156,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 5761
+    },
+    {
+      "epoch": 0.29382429407926136,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 5762
+    },
+    {
+      "epoch": 0.29387528753536674,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 5763
+    },
+    {
+      "epoch": 0.29392628099147217,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 5764
+    },
+    {
+      "epoch": 0.29397727444757754,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 5765
+    },
+    {
+      "epoch": 0.2940282679036829,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 5766
+    },
+    {
+      "epoch": 0.2940792613597883,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 5767
+    },
+    {
+      "epoch": 0.2941302548158937,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 5768
+    },
+    {
+      "epoch": 0.2941812482719991,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 5769
+    },
+    {
+      "epoch": 0.29423224172810447,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2942832351842099,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 5771
+    },
+    {
+      "epoch": 0.2943342286403153,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 5772
+    },
+    {
+      "epoch": 0.29438522209642065,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 5773
+    },
+    {
+      "epoch": 0.294436215552526,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 5774
+    },
+    {
+      "epoch": 0.29448720900863146,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 5775
+    },
+    {
+      "epoch": 0.29453820246473683,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 5776
+    },
+    {
+      "epoch": 0.2945891959208422,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 5777
+    },
+    {
+      "epoch": 0.29464018937694764,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 5778
+    },
+    {
+      "epoch": 0.294691182833053,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8554,
+      "step": 5779
+    },
+    {
+      "epoch": 0.2947421762891584,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 5780
+    },
+    {
+      "epoch": 0.29479316974526376,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 5781
+    },
+    {
+      "epoch": 0.2948441632013692,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 5782
+    },
+    {
+      "epoch": 0.29489515665747457,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 5783
+    },
+    {
+      "epoch": 0.29494615011357994,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 5784
+    },
+    {
+      "epoch": 0.29499714356968537,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5785
+    },
+    {
+      "epoch": 0.29504813702579075,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 5786
+    },
+    {
+      "epoch": 0.2950991304818961,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8773,
+      "step": 5787
+    },
+    {
+      "epoch": 0.2951501239380015,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 5788
+    },
+    {
+      "epoch": 0.2952011173941069,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8734,
+      "step": 5789
+    },
+    {
+      "epoch": 0.2952521108502123,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2953031043063177,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 5791
+    },
+    {
+      "epoch": 0.2953540977624231,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 5792
+    },
+    {
+      "epoch": 0.2954050912185285,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 5793
+    },
+    {
+      "epoch": 0.29545608467463386,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 5794
+    },
+    {
+      "epoch": 0.29550707813073923,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 5795
+    },
+    {
+      "epoch": 0.29555807158684466,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 5796
+    },
+    {
+      "epoch": 0.29560906504295004,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 5797
+    },
+    {
+      "epoch": 0.2956600584990554,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8685,
+      "step": 5798
+    },
+    {
+      "epoch": 0.29571105195516084,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 5799
+    },
+    {
+      "epoch": 0.2957620454112662,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2958130388673716,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 5801
+    },
+    {
+      "epoch": 0.29586403232347697,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5802
+    },
+    {
+      "epoch": 0.2959150257795824,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 5803
+    },
+    {
+      "epoch": 0.29596601923568777,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 5804
+    },
+    {
+      "epoch": 0.29601701269179315,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 5805
+    },
+    {
+      "epoch": 0.2960680061478986,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 5806
+    },
+    {
+      "epoch": 0.29611899960400395,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 5807
+    },
+    {
+      "epoch": 0.2961699930601093,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 5808
+    },
+    {
+      "epoch": 0.2962209865162147,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 5809
+    },
+    {
+      "epoch": 0.29627197997232013,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2963229734284255,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8624,
+      "step": 5811
+    },
+    {
+      "epoch": 0.2963739668845309,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 5812
+    },
+    {
+      "epoch": 0.2964249603406363,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 5813
+    },
+    {
+      "epoch": 0.2964759537967417,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 5814
+    },
+    {
+      "epoch": 0.29652694725284706,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 5815
+    },
+    {
+      "epoch": 0.29657794070895244,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 5816
+    },
+    {
+      "epoch": 0.29662893416505787,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 5817
+    },
+    {
+      "epoch": 0.29667992762116324,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 5818
+    },
+    {
+      "epoch": 0.2967309210772686,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 5819
+    },
+    {
+      "epoch": 0.29678191453337405,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 5820
+    },
+    {
+      "epoch": 0.2968329079894794,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8672,
+      "step": 5821
+    },
+    {
+      "epoch": 0.2968839014455848,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 5822
+    },
+    {
+      "epoch": 0.29693489490169017,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 5823
+    },
+    {
+      "epoch": 0.2969858883577956,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5824
+    },
+    {
+      "epoch": 0.297036881813901,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 5825
+    },
+    {
+      "epoch": 0.29708787527000635,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 5826
+    },
+    {
+      "epoch": 0.2971388687261118,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 5827
+    },
+    {
+      "epoch": 0.29718986218221716,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 5828
+    },
+    {
+      "epoch": 0.29724085563832253,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 5829
+    },
+    {
+      "epoch": 0.2972918490944279,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.843,
+      "step": 5830
+    },
+    {
+      "epoch": 0.29734284255053334,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 5831
+    },
+    {
+      "epoch": 0.2973938360066387,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 5832
+    },
+    {
+      "epoch": 0.2974448294627441,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 5833
+    },
+    {
+      "epoch": 0.2974958229188495,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 5834
+    },
+    {
+      "epoch": 0.2975468163749549,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 5835
+    },
+    {
+      "epoch": 0.29759780983106027,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 5836
+    },
+    {
+      "epoch": 0.29764880328716564,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 5837
+    },
+    {
+      "epoch": 0.2976997967432711,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 5838
+    },
+    {
+      "epoch": 0.29775079019937645,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 5839
+    },
+    {
+      "epoch": 0.2978017836554818,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 5840
+    },
+    {
+      "epoch": 0.29785277711158725,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 5841
+    },
+    {
+      "epoch": 0.2979037705676926,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 5842
+    },
+    {
+      "epoch": 0.297954764023798,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 5843
+    },
+    {
+      "epoch": 0.2980057574799034,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 5844
+    },
+    {
+      "epoch": 0.2980567509360088,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2981077443921142,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 5846
+    },
+    {
+      "epoch": 0.29815873784821956,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 5847
+    },
+    {
+      "epoch": 0.298209731304325,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 5848
+    },
+    {
+      "epoch": 0.29826072476043036,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 5849
+    },
+    {
+      "epoch": 0.29831171821653574,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2983627116726411,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 5851
+    },
+    {
+      "epoch": 0.29841370512874654,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 5852
+    },
+    {
+      "epoch": 0.2984646985848519,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 5853
+    },
+    {
+      "epoch": 0.2985156920409573,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 5854
+    },
+    {
+      "epoch": 0.2985666854970627,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 5855
+    },
+    {
+      "epoch": 0.2986176789531681,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 5856
+    },
+    {
+      "epoch": 0.2986686724092735,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5857
+    },
+    {
+      "epoch": 0.29871966586537885,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 5858
+    },
+    {
+      "epoch": 0.2987706593214843,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8725,
+      "step": 5859
+    },
+    {
+      "epoch": 0.29882165277758965,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5860
+    },
+    {
+      "epoch": 0.298872646233695,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 5861
+    },
+    {
+      "epoch": 0.29892363968980046,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 5862
+    },
+    {
+      "epoch": 0.29897463314590583,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 5863
+    },
+    {
+      "epoch": 0.2990256266020112,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 5864
+    },
+    {
+      "epoch": 0.2990766200581166,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 5865
+    },
+    {
+      "epoch": 0.299127613514222,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2991786069703274,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8747,
+      "step": 5867
+    },
+    {
+      "epoch": 0.29922960042643276,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 5868
+    },
+    {
+      "epoch": 0.2992805938825382,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8589,
+      "step": 5869
+    },
+    {
+      "epoch": 0.29933158733864357,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5870
+    },
+    {
+      "epoch": 0.29938258079474894,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 5871
+    },
+    {
+      "epoch": 0.2994335742508543,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 5872
+    },
+    {
+      "epoch": 0.29948456770695975,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 5873
+    },
+    {
+      "epoch": 0.2995355611630651,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 5874
+    },
+    {
+      "epoch": 0.2995865546191705,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 5875
+    },
+    {
+      "epoch": 0.29963754807527593,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 5876
+    },
+    {
+      "epoch": 0.2996885415313813,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 5877
+    },
+    {
+      "epoch": 0.2997395349874867,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 5878
+    },
+    {
+      "epoch": 0.29979052844359205,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8603,
+      "step": 5879
+    },
+    {
+      "epoch": 0.2998415218996975,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 5880
+    },
+    {
+      "epoch": 0.29989251535580286,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 5881
+    },
+    {
+      "epoch": 0.29994350881190823,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 5882
+    },
+    {
+      "epoch": 0.29999450226801366,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 5883
+    },
+    {
+      "epoch": 0.30004549572411904,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8691,
+      "step": 5884
+    },
+    {
+      "epoch": 0.3000964891802244,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3001474826363298,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 5886
+    },
+    {
+      "epoch": 0.3001984760924352,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 5887
+    },
+    {
+      "epoch": 0.3002494695485406,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 5888
+    },
+    {
+      "epoch": 0.30030046300464597,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 5889
+    },
+    {
+      "epoch": 0.3003514564607514,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3004024499168568,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 5891
+    },
+    {
+      "epoch": 0.30045344337296215,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 5892
+    },
+    {
+      "epoch": 0.3005044368290675,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 5893
+    },
+    {
+      "epoch": 0.30055543028517295,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 5894
+    },
+    {
+      "epoch": 0.30060642374127833,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3006574171973837,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 5896
+    },
+    {
+      "epoch": 0.30070841065348913,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 5897
+    },
+    {
+      "epoch": 0.3007594041095945,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 5898
+    },
+    {
+      "epoch": 0.3008103975656999,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 5899
+    },
+    {
+      "epoch": 0.30086139102180526,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3009123844779107,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8697,
+      "step": 5901
+    },
+    {
+      "epoch": 0.30096337793401606,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 5902
+    },
+    {
+      "epoch": 0.30101437139012144,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 5903
+    },
+    {
+      "epoch": 0.30106536484622687,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 5904
+    },
+    {
+      "epoch": 0.30111635830233224,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 5905
+    },
+    {
+      "epoch": 0.3011673517584376,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 5906
+    },
+    {
+      "epoch": 0.301218345214543,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5907
+    },
+    {
+      "epoch": 0.3012693386706484,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 5908
+    },
+    {
+      "epoch": 0.3013203321267538,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8593,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3013713255828592,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3014223190389646,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 5911
+    },
+    {
+      "epoch": 0.30147331249507,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 5912
+    },
+    {
+      "epoch": 0.30152430595117535,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 5913
+    },
+    {
+      "epoch": 0.30157529940728073,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 5914
+    },
+    {
+      "epoch": 0.30162629286338616,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8802,
+      "step": 5915
+    },
+    {
+      "epoch": 0.30167728631949153,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3017282797755969,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8786,
+      "step": 5917
+    },
+    {
+      "epoch": 0.30177927323170234,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 5918
+    },
+    {
+      "epoch": 0.3018302666878077,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 5919
+    },
+    {
+      "epoch": 0.3018812601439131,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 5920
+    },
+    {
+      "epoch": 0.30193225360001846,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8598,
+      "step": 5921
+    },
+    {
+      "epoch": 0.3019832470561239,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8541,
+      "step": 5922
+    },
+    {
+      "epoch": 0.30203424051222927,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 5923
+    },
+    {
+      "epoch": 0.30208523396833464,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3021362274244401,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 5925
+    },
+    {
+      "epoch": 0.30218722088054545,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8601,
+      "step": 5926
+    },
+    {
+      "epoch": 0.3022382143366508,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 5927
+    },
+    {
+      "epoch": 0.3022892077927562,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 5928
+    },
+    {
+      "epoch": 0.30234020124886163,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 5929
+    },
+    {
+      "epoch": 0.302391194704967,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3024421881610724,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 5931
+    },
+    {
+      "epoch": 0.3024931816171778,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 5932
+    },
+    {
+      "epoch": 0.3025441750732832,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 5933
+    },
+    {
+      "epoch": 0.30259516852938856,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 5934
+    },
+    {
+      "epoch": 0.30264616198549393,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 5935
+    },
+    {
+      "epoch": 0.30269715544159936,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 5936
+    },
+    {
+      "epoch": 0.30274814889770474,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8587,
+      "step": 5937
+    },
+    {
+      "epoch": 0.3027991423538101,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 5938
+    },
+    {
+      "epoch": 0.30285013580991554,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 5939
+    },
+    {
+      "epoch": 0.3029011292660209,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3029521227221263,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 5941
+    },
+    {
+      "epoch": 0.30300311617823167,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 5942
+    },
+    {
+      "epoch": 0.3030541096343371,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 5943
+    },
+    {
+      "epoch": 0.3031051030904425,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 5944
+    },
+    {
+      "epoch": 0.30315609654654785,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8564,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3032070900026533,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 5946
+    },
+    {
+      "epoch": 0.30325808345875865,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 5947
+    },
+    {
+      "epoch": 0.30330907691486403,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8479,
+      "step": 5948
+    },
+    {
+      "epoch": 0.3033600703709694,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 5949
+    },
+    {
+      "epoch": 0.30341106382707483,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3034620572831802,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.849,
+      "step": 5951
+    },
+    {
+      "epoch": 0.3035130507392856,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 5952
+    },
+    {
+      "epoch": 0.303564044195391,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 5953
+    },
+    {
+      "epoch": 0.3036150376514964,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 5954
+    },
+    {
+      "epoch": 0.30366603110760176,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 5955
+    },
+    {
+      "epoch": 0.30371702456370714,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 5956
+    },
+    {
+      "epoch": 0.30376801801981257,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 5957
+    },
+    {
+      "epoch": 0.30381901147591794,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8687,
+      "step": 5958
+    },
+    {
+      "epoch": 0.3038700049320233,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 5959
+    },
+    {
+      "epoch": 0.30392099838812875,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3039719918442341,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 5961
+    },
+    {
+      "epoch": 0.3040229853003395,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.868,
+      "step": 5962
+    },
+    {
+      "epoch": 0.3040739787564449,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 5963
+    },
+    {
+      "epoch": 0.3041249722125503,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 5964
+    },
+    {
+      "epoch": 0.3041759656686557,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 5965
+    },
+    {
+      "epoch": 0.30422695912476105,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 5966
+    },
+    {
+      "epoch": 0.3042779525808665,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 5967
+    },
+    {
+      "epoch": 0.30432894603697186,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 5968
+    },
+    {
+      "epoch": 0.30437993949307723,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 5969
+    },
+    {
+      "epoch": 0.3044309329491826,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 5970
+    },
+    {
+      "epoch": 0.30448192640528804,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 5971
+    },
+    {
+      "epoch": 0.3045329198613934,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 5972
+    },
+    {
+      "epoch": 0.3045839133174988,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 5973
+    },
+    {
+      "epoch": 0.3046349067736042,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 5974
+    },
+    {
+      "epoch": 0.3046859002297096,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 5975
+    },
+    {
+      "epoch": 0.30473689368581497,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8789,
+      "step": 5976
+    },
+    {
+      "epoch": 0.30478788714192034,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 5977
+    },
+    {
+      "epoch": 0.3048388805980258,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 5978
+    },
+    {
+      "epoch": 0.30488987405413115,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 5979
+    },
+    {
+      "epoch": 0.3049408675102365,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 5980
+    },
+    {
+      "epoch": 0.30499186096634195,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 5981
+    },
+    {
+      "epoch": 0.30504285442244733,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 5982
+    },
+    {
+      "epoch": 0.3050938478785527,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 5983
+    },
+    {
+      "epoch": 0.3051448413346581,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8626,
+      "step": 5984
+    },
+    {
+      "epoch": 0.3051958347907635,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 5985
+    },
+    {
+      "epoch": 0.3052468282468689,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 5986
+    },
+    {
+      "epoch": 0.30529782170297426,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 5987
+    },
+    {
+      "epoch": 0.3053488151590797,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 5988
+    },
+    {
+      "epoch": 0.30539980861518506,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8487,
+      "step": 5989
+    },
+    {
+      "epoch": 0.30545080207129044,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3055017955273958,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 5991
+    },
+    {
+      "epoch": 0.30555278898350124,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 5992
+    },
+    {
+      "epoch": 0.3056037824396066,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 5993
+    },
+    {
+      "epoch": 0.305654775895712,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 5994
+    },
+    {
+      "epoch": 0.3057057693518174,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 5995
+    },
+    {
+      "epoch": 0.3057567628079228,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8652,
+      "step": 5996
+    },
+    {
+      "epoch": 0.3058077562640282,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 5997
+    },
+    {
+      "epoch": 0.30585874972013355,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 5998
+    },
+    {
+      "epoch": 0.305909743176239,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 5999
+    },
+    {
+      "epoch": 0.30596073663234435,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 6000
+    },
+    {
+      "epoch": 0.30601173008844973,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8445,
+      "step": 6001
+    },
+    {
+      "epoch": 0.30606272354455516,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 6002
+    },
+    {
+      "epoch": 0.30611371700066053,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 6003
+    },
+    {
+      "epoch": 0.3061647104567659,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 6004
+    },
+    {
+      "epoch": 0.3062157039128713,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 6005
+    },
+    {
+      "epoch": 0.3062666973689767,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 6006
+    },
+    {
+      "epoch": 0.3063176908250821,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 6007
+    },
+    {
+      "epoch": 0.30636868428118746,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 6008
+    },
+    {
+      "epoch": 0.3064196777372929,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 6009
+    },
+    {
+      "epoch": 0.30647067119339827,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 6010
+    },
+    {
+      "epoch": 0.30652166464950364,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 6011
+    },
+    {
+      "epoch": 0.306572658105609,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 6012
+    },
+    {
+      "epoch": 0.30662365156171445,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 6013
+    },
+    {
+      "epoch": 0.3066746450178198,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 6014
+    },
+    {
+      "epoch": 0.3067256384739252,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6015
+    },
+    {
+      "epoch": 0.30677663193003063,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 6016
+    },
+    {
+      "epoch": 0.306827625386136,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 6017
+    },
+    {
+      "epoch": 0.3068786188422414,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 6018
+    },
+    {
+      "epoch": 0.30692961229834675,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 6019
+    },
+    {
+      "epoch": 0.3069806057544522,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 6020
+    },
+    {
+      "epoch": 0.30703159921055756,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 6021
+    },
+    {
+      "epoch": 0.30708259266666293,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 6022
+    },
+    {
+      "epoch": 0.30713358612276837,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 6023
+    },
+    {
+      "epoch": 0.30718457957887374,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 6024
+    },
+    {
+      "epoch": 0.3072355730349791,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 6025
+    },
+    {
+      "epoch": 0.3072865664910845,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 6026
+    },
+    {
+      "epoch": 0.3073375599471899,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 6027
+    },
+    {
+      "epoch": 0.3073885534032953,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 6028
+    },
+    {
+      "epoch": 0.30743954685940067,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 6029
+    },
+    {
+      "epoch": 0.3074905403155061,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3075415337716115,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 6031
+    },
+    {
+      "epoch": 0.30759252722771685,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6032
+    },
+    {
+      "epoch": 0.3076435206838222,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 6033
+    },
+    {
+      "epoch": 0.30769451413992766,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 6034
+    },
+    {
+      "epoch": 0.30774550759603303,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 6035
+    },
+    {
+      "epoch": 0.3077965010521384,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.86,
+      "step": 6036
+    },
+    {
+      "epoch": 0.30784749450824384,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 6037
+    },
+    {
+      "epoch": 0.3078984879643492,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 6038
+    },
+    {
+      "epoch": 0.3079494814204546,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 6039
+    },
+    {
+      "epoch": 0.30800047487655996,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 6040
+    },
+    {
+      "epoch": 0.3080514683326654,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 6041
+    },
+    {
+      "epoch": 0.30810246178877077,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 6042
+    },
+    {
+      "epoch": 0.30815345524487614,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 6043
+    },
+    {
+      "epoch": 0.30820444870098157,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 6044
+    },
+    {
+      "epoch": 0.30825544215708695,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6045
+    },
+    {
+      "epoch": 0.3083064356131923,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8623,
+      "step": 6046
+    },
+    {
+      "epoch": 0.3083574290692977,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6047
+    },
+    {
+      "epoch": 0.3084084225254031,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 6048
+    },
+    {
+      "epoch": 0.3084594159815085,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6049
+    },
+    {
+      "epoch": 0.3085104094376139,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.866,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3085614028937193,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 6051
+    },
+    {
+      "epoch": 0.3086123963498247,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 6052
+    },
+    {
+      "epoch": 0.30866338980593006,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 6053
+    },
+    {
+      "epoch": 0.30871438326203543,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 6054
+    },
+    {
+      "epoch": 0.30876537671814086,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 6055
+    },
+    {
+      "epoch": 0.30881637017424624,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6056
+    },
+    {
+      "epoch": 0.3088673636303516,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 6057
+    },
+    {
+      "epoch": 0.30891835708645704,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 6058
+    },
+    {
+      "epoch": 0.3089693505425624,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 6059
+    },
+    {
+      "epoch": 0.3090203439986678,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 6060
+    },
+    {
+      "epoch": 0.30907133745477317,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 6061
+    },
+    {
+      "epoch": 0.3091223309108786,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 6062
+    },
+    {
+      "epoch": 0.30917332436698397,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6063
+    },
+    {
+      "epoch": 0.30922431782308935,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 6064
+    },
+    {
+      "epoch": 0.3092753112791948,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 6065
+    },
+    {
+      "epoch": 0.30932630473530015,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6066
+    },
+    {
+      "epoch": 0.3093772981914055,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 6067
+    },
+    {
+      "epoch": 0.3094282916475109,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 6068
+    },
+    {
+      "epoch": 0.30947928510361633,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 6069
+    },
+    {
+      "epoch": 0.3095302785597217,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 6070
+    },
+    {
+      "epoch": 0.3095812720158271,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 6071
+    },
+    {
+      "epoch": 0.3096322654719325,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 6072
+    },
+    {
+      "epoch": 0.3096832589280379,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 6073
+    },
+    {
+      "epoch": 0.30973425238414326,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 6074
+    },
+    {
+      "epoch": 0.30978524584024864,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8683,
+      "step": 6075
+    },
+    {
+      "epoch": 0.30983623929635407,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6076
+    },
+    {
+      "epoch": 0.30988723275245944,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 6077
+    },
+    {
+      "epoch": 0.3099382262085648,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6078
+    },
+    {
+      "epoch": 0.30998921966467025,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 6079
+    },
+    {
+      "epoch": 0.3100402131207756,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8521,
+      "step": 6080
+    },
+    {
+      "epoch": 0.310091206576881,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 6081
+    },
+    {
+      "epoch": 0.31014220003298637,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 6082
+    },
+    {
+      "epoch": 0.3101931934890918,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 6083
+    },
+    {
+      "epoch": 0.3102441869451972,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 6084
+    },
+    {
+      "epoch": 0.31029518040130255,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 6085
+    },
+    {
+      "epoch": 0.310346173857408,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 6086
+    },
+    {
+      "epoch": 0.31039716731351336,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 6087
+    },
+    {
+      "epoch": 0.31044816076961873,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 6088
+    },
+    {
+      "epoch": 0.3104991542257241,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 6089
+    },
+    {
+      "epoch": 0.31055014768182954,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3106011411379349,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 6091
+    },
+    {
+      "epoch": 0.3106521345940403,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 6092
+    },
+    {
+      "epoch": 0.3107031280501457,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 6093
+    },
+    {
+      "epoch": 0.3107541215062511,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 6094
+    },
+    {
+      "epoch": 0.31080511496235647,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 6095
+    },
+    {
+      "epoch": 0.31085610841846184,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 6096
+    },
+    {
+      "epoch": 0.31090710187456727,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 6097
+    },
+    {
+      "epoch": 0.31095809533067265,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 6098
+    },
+    {
+      "epoch": 0.311009088786778,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 6099
+    },
+    {
+      "epoch": 0.31106008224288345,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3111110756989888,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 6101
+    },
+    {
+      "epoch": 0.3111620691550942,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 6102
+    },
+    {
+      "epoch": 0.3112130626111996,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 6103
+    },
+    {
+      "epoch": 0.311264056067305,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 6104
+    },
+    {
+      "epoch": 0.3113150495234104,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 6105
+    },
+    {
+      "epoch": 0.31136604297951576,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 6106
+    },
+    {
+      "epoch": 0.3114170364356212,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 6107
+    },
+    {
+      "epoch": 0.31146802989172656,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 6108
+    },
+    {
+      "epoch": 0.31151902334783194,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8609,
+      "step": 6109
+    },
+    {
+      "epoch": 0.3115700168039373,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8579,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31162101026004274,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 6111
+    },
+    {
+      "epoch": 0.3116720037161481,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 6112
+    },
+    {
+      "epoch": 0.3117229971722535,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 6113
+    },
+    {
+      "epoch": 0.3117739906283589,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 6114
+    },
+    {
+      "epoch": 0.3118249840844643,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 6115
+    },
+    {
+      "epoch": 0.31187597754056967,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 6116
+    },
+    {
+      "epoch": 0.31192697099667505,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 6117
+    },
+    {
+      "epoch": 0.3119779644527805,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 6118
+    },
+    {
+      "epoch": 0.31202895790888585,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 6119
+    },
+    {
+      "epoch": 0.3120799513649912,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31213094482109666,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 6121
+    },
+    {
+      "epoch": 0.31218193827720203,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 6122
+    },
+    {
+      "epoch": 0.3122329317333074,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 6123
+    },
+    {
+      "epoch": 0.3122839251894128,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 6124
+    },
+    {
+      "epoch": 0.3123349186455182,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 6125
+    },
+    {
+      "epoch": 0.3123859121016236,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 6126
+    },
+    {
+      "epoch": 0.31243690555772896,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6127
+    },
+    {
+      "epoch": 0.3124878990138344,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 6128
+    },
+    {
+      "epoch": 0.31253889246993977,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 6129
+    },
+    {
+      "epoch": 0.31258988592604514,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 6130
+    },
+    {
+      "epoch": 0.3126408793821505,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 6131
+    },
+    {
+      "epoch": 0.31269187283825595,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 6132
+    },
+    {
+      "epoch": 0.3127428662943613,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 6133
+    },
+    {
+      "epoch": 0.3127938597504667,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 6134
+    },
+    {
+      "epoch": 0.3128448532065721,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 6135
+    },
+    {
+      "epoch": 0.3128958466626775,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 6136
+    },
+    {
+      "epoch": 0.3129468401187829,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.846,
+      "step": 6137
+    },
+    {
+      "epoch": 0.31299783357488825,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 6138
+    },
+    {
+      "epoch": 0.3130488270309937,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 6139
+    },
+    {
+      "epoch": 0.31309982048709906,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 6140
+    },
+    {
+      "epoch": 0.31315081394320443,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8695,
+      "step": 6141
+    },
+    {
+      "epoch": 0.31320180739930986,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 6142
+    },
+    {
+      "epoch": 0.31325280085541524,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6143
+    },
+    {
+      "epoch": 0.3133037943115206,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 6144
+    },
+    {
+      "epoch": 0.313354787767626,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 6145
+    },
+    {
+      "epoch": 0.3134057812237314,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 6146
+    },
+    {
+      "epoch": 0.3134567746798368,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 6147
+    },
+    {
+      "epoch": 0.31350776813594217,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 6148
+    },
+    {
+      "epoch": 0.3135587615920476,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 6149
+    },
+    {
+      "epoch": 0.31360975504815297,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 6150
+    },
+    {
+      "epoch": 0.31366074850425835,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 6151
+    },
+    {
+      "epoch": 0.3137117419603637,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 6152
+    },
+    {
+      "epoch": 0.31376273541646915,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8702,
+      "step": 6153
+    },
+    {
+      "epoch": 0.3138137288725745,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 6154
+    },
+    {
+      "epoch": 0.3138647223286799,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 6155
+    },
+    {
+      "epoch": 0.31391571578478533,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 6156
+    },
+    {
+      "epoch": 0.3139667092408907,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 6157
+    },
+    {
+      "epoch": 0.3140177026969961,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8553,
+      "step": 6158
+    },
+    {
+      "epoch": 0.31406869615310146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 6159
+    },
+    {
+      "epoch": 0.3141196896092069,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 6160
+    },
+    {
+      "epoch": 0.31417068306531226,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8597,
+      "step": 6161
+    },
+    {
+      "epoch": 0.31422167652141764,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 6162
+    },
+    {
+      "epoch": 0.31427266997752307,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 6163
+    },
+    {
+      "epoch": 0.31432366343362844,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 6164
+    },
+    {
+      "epoch": 0.3143746568897338,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6165
+    },
+    {
+      "epoch": 0.3144256503458392,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 6166
+    },
+    {
+      "epoch": 0.3144766438019446,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 6167
+    },
+    {
+      "epoch": 0.31452763725805,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 6168
+    },
+    {
+      "epoch": 0.31457863071415537,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 6169
+    },
+    {
+      "epoch": 0.3146296241702608,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8433,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3146806176263662,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 6171
+    },
+    {
+      "epoch": 0.31473161108247155,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 6172
+    },
+    {
+      "epoch": 0.3147826045385769,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 6173
+    },
+    {
+      "epoch": 0.31483359799468236,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 6174
+    },
+    {
+      "epoch": 0.31488459145078773,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 6175
+    },
+    {
+      "epoch": 0.3149355849068931,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 6176
+    },
+    {
+      "epoch": 0.31498657836299854,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 6177
+    },
+    {
+      "epoch": 0.3150375718191039,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 6178
+    },
+    {
+      "epoch": 0.3150885652752093,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 6179
+    },
+    {
+      "epoch": 0.31513955873131466,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 6180
+    },
+    {
+      "epoch": 0.3151905521874201,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 6181
+    },
+    {
+      "epoch": 0.31524154564352547,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 6182
+    },
+    {
+      "epoch": 0.31529253909963084,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.002,
+      "loss": 2.8525,
+      "step": 6183
+    },
+    {
+      "epoch": 0.3153435325557363,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 6184
+    },
+    {
+      "epoch": 0.31539452601184165,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 6185
+    },
+    {
+      "epoch": 0.315445519467947,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 6186
+    },
+    {
+      "epoch": 0.3154965129240524,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3155475063801578,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3155984998362632,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 6189
+    },
+    {
+      "epoch": 0.3156494932923686,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 6190
+    },
+    {
+      "epoch": 0.315700486748474,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3157514802045794,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 6192
+    },
+    {
+      "epoch": 0.31580247366068476,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 6193
+    },
+    {
+      "epoch": 0.31585346711679013,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.863,
+      "step": 6194
+    },
+    {
+      "epoch": 0.31590446057289556,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 6195
+    },
+    {
+      "epoch": 0.31595545402900094,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3160064474851063,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 6197
+    },
+    {
+      "epoch": 0.31605744094121174,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 6198
+    },
+    {
+      "epoch": 0.3161084343973171,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 6199
+    },
+    {
+      "epoch": 0.3161594278534225,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 6200
+    },
+    {
+      "epoch": 0.31621042130952787,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 6201
+    },
+    {
+      "epoch": 0.3162614147656333,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 6202
+    },
+    {
+      "epoch": 0.3163124082217387,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 6203
+    },
+    {
+      "epoch": 0.31636340167784405,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 6204
+    },
+    {
+      "epoch": 0.3164143951339495,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 6205
+    },
+    {
+      "epoch": 0.31646538859005485,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 6206
+    },
+    {
+      "epoch": 0.3165163820461602,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 6207
+    },
+    {
+      "epoch": 0.3165673755022656,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 6208
+    },
+    {
+      "epoch": 0.31661836895837103,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 6209
+    },
+    {
+      "epoch": 0.3166693624144764,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3167203558705818,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 6211
+    },
+    {
+      "epoch": 0.3167713493266872,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 6212
+    },
+    {
+      "epoch": 0.3168223427827926,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 6213
+    },
+    {
+      "epoch": 0.31687333623889796,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 6214
+    },
+    {
+      "epoch": 0.31692432969500334,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 6215
+    },
+    {
+      "epoch": 0.31697532315110877,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 6216
+    },
+    {
+      "epoch": 0.31702631660721414,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 6217
+    },
+    {
+      "epoch": 0.3170773100633195,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 6218
+    },
+    {
+      "epoch": 0.31712830351942495,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8658,
+      "step": 6219
+    },
+    {
+      "epoch": 0.3171792969755303,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8628,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3172302904316357,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 6221
+    },
+    {
+      "epoch": 0.3172812838877411,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 6222
+    },
+    {
+      "epoch": 0.3173322773438465,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 6223
+    },
+    {
+      "epoch": 0.3173832707999519,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 6224
+    },
+    {
+      "epoch": 0.31743426425605725,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 6225
+    },
+    {
+      "epoch": 0.3174852577121627,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 6226
+    },
+    {
+      "epoch": 0.31753625116826806,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 6227
+    },
+    {
+      "epoch": 0.31758724462437343,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 6228
+    },
+    {
+      "epoch": 0.3176382380804788,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 6229
+    },
+    {
+      "epoch": 0.31768923153658424,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3177402249926896,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8588,
+      "step": 6231
+    },
+    {
+      "epoch": 0.317791218448795,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 6232
+    },
+    {
+      "epoch": 0.3178422119049004,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6233
+    },
+    {
+      "epoch": 0.3178932053610058,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 6234
+    },
+    {
+      "epoch": 0.31794419881711117,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 6235
+    },
+    {
+      "epoch": 0.31799519227321654,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8604,
+      "step": 6236
+    },
+    {
+      "epoch": 0.318046185729322,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 6237
+    },
+    {
+      "epoch": 0.31809717918542735,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 6238
+    },
+    {
+      "epoch": 0.3181481726415327,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 6239
+    },
+    {
+      "epoch": 0.31819916609763815,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 6240
+    },
+    {
+      "epoch": 0.31825015955374353,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 6241
+    },
+    {
+      "epoch": 0.3183011530098489,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 6242
+    },
+    {
+      "epoch": 0.3183521464659543,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 6243
+    },
+    {
+      "epoch": 0.3184031399220597,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 6244
+    },
+    {
+      "epoch": 0.3184541333781651,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8754,
+      "step": 6245
+    },
+    {
+      "epoch": 0.31850512683427046,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 6246
+    },
+    {
+      "epoch": 0.3185561202903759,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 6247
+    },
+    {
+      "epoch": 0.31860711374648126,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 6248
+    },
+    {
+      "epoch": 0.31865810720258664,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8588,
+      "step": 6249
+    },
+    {
+      "epoch": 0.318709100658692,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 6250
+    },
+    {
+      "epoch": 0.31876009411479744,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 6251
+    },
+    {
+      "epoch": 0.3188110875709028,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 6252
+    },
+    {
+      "epoch": 0.3188620810270082,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 6253
+    },
+    {
+      "epoch": 0.3189130744831136,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 6254
+    },
+    {
+      "epoch": 0.318964067939219,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 6255
+    },
+    {
+      "epoch": 0.3190150613953244,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 6256
+    },
+    {
+      "epoch": 0.31906605485142975,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 6257
+    },
+    {
+      "epoch": 0.3191170483075352,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 6258
+    },
+    {
+      "epoch": 0.31916804176364055,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6259
+    },
+    {
+      "epoch": 0.31921903521974593,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31927002867585136,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 6261
+    },
+    {
+      "epoch": 0.31932102213195673,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 6262
+    },
+    {
+      "epoch": 0.3193720155880621,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 6263
+    },
+    {
+      "epoch": 0.3194230090441675,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 6264
+    },
+    {
+      "epoch": 0.3194740025002729,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 6265
+    },
+    {
+      "epoch": 0.3195249959563783,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 6266
+    },
+    {
+      "epoch": 0.31957598941248366,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 6267
+    },
+    {
+      "epoch": 0.3196269828685891,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.866,
+      "step": 6268
+    },
+    {
+      "epoch": 0.31967797632469447,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 6269
+    },
+    {
+      "epoch": 0.31972896978079984,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6270
+    },
+    {
+      "epoch": 0.3197799632369052,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 6271
+    },
+    {
+      "epoch": 0.31983095669301065,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 6272
+    },
+    {
+      "epoch": 0.319881950149116,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 6273
+    },
+    {
+      "epoch": 0.3199329436052214,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 6274
+    },
+    {
+      "epoch": 0.31998393706132683,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 6275
+    },
+    {
+      "epoch": 0.3200349305174322,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 6276
+    },
+    {
+      "epoch": 0.3200859239735376,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 6277
+    },
+    {
+      "epoch": 0.32013691742964295,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 6278
+    },
+    {
+      "epoch": 0.3201879108857484,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6279
+    },
+    {
+      "epoch": 0.32023890434185376,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 6280
+    },
+    {
+      "epoch": 0.32028989779795913,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 6281
+    },
+    {
+      "epoch": 0.32034089125406456,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 6282
+    },
+    {
+      "epoch": 0.32039188471016994,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 6283
+    },
+    {
+      "epoch": 0.3204428781662753,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 6284
+    },
+    {
+      "epoch": 0.3204938716223807,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 6285
+    },
+    {
+      "epoch": 0.3205448650784861,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 6286
+    },
+    {
+      "epoch": 0.3205958585345915,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 6287
+    },
+    {
+      "epoch": 0.32064685199069687,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 6288
+    },
+    {
+      "epoch": 0.3206978454468023,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8631,
+      "step": 6289
+    },
+    {
+      "epoch": 0.3207488389029077,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 6290
+    },
+    {
+      "epoch": 0.32079983235901305,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 6291
+    },
+    {
+      "epoch": 0.3208508258151184,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 6292
+    },
+    {
+      "epoch": 0.32090181927122385,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 6293
+    },
+    {
+      "epoch": 0.32095281272732923,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 6294
+    },
+    {
+      "epoch": 0.3210038061834346,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 6295
+    },
+    {
+      "epoch": 0.32105479963954003,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 6296
+    },
+    {
+      "epoch": 0.3211057930956454,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 6297
+    },
+    {
+      "epoch": 0.3211567865517508,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 6298
+    },
+    {
+      "epoch": 0.32120778000785616,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6299
+    },
+    {
+      "epoch": 0.3212587734639616,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 6300
+    },
+    {
+      "epoch": 0.32130976692006696,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 6301
+    },
+    {
+      "epoch": 0.32136076037617234,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 6302
+    },
+    {
+      "epoch": 0.32141175383227777,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 6303
+    },
+    {
+      "epoch": 0.32146274728838314,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 6304
+    },
+    {
+      "epoch": 0.3215137407444885,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 6305
+    },
+    {
+      "epoch": 0.3215647342005939,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 6306
+    },
+    {
+      "epoch": 0.3216157276566993,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8706,
+      "step": 6307
+    },
+    {
+      "epoch": 0.3216667211128047,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8605,
+      "step": 6308
+    },
+    {
+      "epoch": 0.3217177145689101,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 6309
+    },
+    {
+      "epoch": 0.3217687080250155,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 6310
+    },
+    {
+      "epoch": 0.3218197014811209,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 6311
+    },
+    {
+      "epoch": 0.32187069493722625,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 6312
+    },
+    {
+      "epoch": 0.32192168839333163,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 6313
+    },
+    {
+      "epoch": 0.32197268184943706,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 6314
+    },
+    {
+      "epoch": 0.32202367530554243,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 6315
+    },
+    {
+      "epoch": 0.3220746687616478,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8678,
+      "step": 6316
+    },
+    {
+      "epoch": 0.32212566221775324,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 6317
+    },
+    {
+      "epoch": 0.3221766556738586,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 6318
+    },
+    {
+      "epoch": 0.322227649129964,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 6319
+    },
+    {
+      "epoch": 0.32227864258606936,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 6320
+    },
+    {
+      "epoch": 0.3223296360421748,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 6321
+    },
+    {
+      "epoch": 0.32238062949828017,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 6322
+    },
+    {
+      "epoch": 0.32243162295438554,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 6323
+    },
+    {
+      "epoch": 0.322482616410491,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 6324
+    },
+    {
+      "epoch": 0.32253360986659635,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 6325
+    },
+    {
+      "epoch": 0.3225846033227017,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 6326
+    },
+    {
+      "epoch": 0.3226355967788071,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 6327
+    },
+    {
+      "epoch": 0.32268659023491253,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 6328
+    },
+    {
+      "epoch": 0.3227375836910179,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 6329
+    },
+    {
+      "epoch": 0.3227885771471233,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3228395706032287,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 6331
+    },
+    {
+      "epoch": 0.3228905640593341,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 6332
+    },
+    {
+      "epoch": 0.32294155751543946,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 6333
+    },
+    {
+      "epoch": 0.32299255097154483,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 6334
+    },
+    {
+      "epoch": 0.32304354442765026,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 6335
+    },
+    {
+      "epoch": 0.32309453788375564,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 6336
+    },
+    {
+      "epoch": 0.323145531339861,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 6337
+    },
+    {
+      "epoch": 0.32319652479596644,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 6338
+    },
+    {
+      "epoch": 0.3232475182520718,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 6339
+    },
+    {
+      "epoch": 0.3232985117081772,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 6340
+    },
+    {
+      "epoch": 0.32334950516428257,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 6341
+    },
+    {
+      "epoch": 0.323400498620388,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 6342
+    },
+    {
+      "epoch": 0.3234514920764934,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 6343
+    },
+    {
+      "epoch": 0.32350248553259875,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8562,
+      "step": 6344
+    },
+    {
+      "epoch": 0.3235534789887042,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 6345
+    },
+    {
+      "epoch": 0.32360447244480955,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 6346
+    },
+    {
+      "epoch": 0.32365546590091493,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 6347
+    },
+    {
+      "epoch": 0.3237064593570203,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 6348
+    },
+    {
+      "epoch": 0.32375745281312573,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 6349
+    },
+    {
+      "epoch": 0.3238084462692311,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3238594397253365,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 6351
+    },
+    {
+      "epoch": 0.3239104331814419,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 6352
+    },
+    {
+      "epoch": 0.3239614266375473,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 6353
+    },
+    {
+      "epoch": 0.32401242009365266,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 6354
+    },
+    {
+      "epoch": 0.32406341354975804,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8549,
+      "step": 6355
+    },
+    {
+      "epoch": 0.32411440700586347,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 6356
+    },
+    {
+      "epoch": 0.32416540046196884,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 6357
+    },
+    {
+      "epoch": 0.3242163939180742,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 6358
+    },
+    {
+      "epoch": 0.32426738737417965,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 6359
+    },
+    {
+      "epoch": 0.324318380830285,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3243693742863904,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 6361
+    },
+    {
+      "epoch": 0.3244203677424958,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 6362
+    },
+    {
+      "epoch": 0.3244713611986012,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 6363
+    },
+    {
+      "epoch": 0.3245223546547066,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 6364
+    },
+    {
+      "epoch": 0.32457334811081195,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8655,
+      "step": 6365
+    },
+    {
+      "epoch": 0.3246243415669174,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 6366
+    },
+    {
+      "epoch": 0.32467533502302276,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 6367
+    },
+    {
+      "epoch": 0.32472632847912813,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6368
+    },
+    {
+      "epoch": 0.3247773219352335,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 6369
+    },
+    {
+      "epoch": 0.32482831539133894,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3248793088474443,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 6371
+    },
+    {
+      "epoch": 0.3249303023035497,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3249812957596551,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 6373
+    },
+    {
+      "epoch": 0.3250322892157605,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 6374
+    },
+    {
+      "epoch": 0.32508328267186587,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 6375
+    },
+    {
+      "epoch": 0.32513427612797124,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 6376
+    },
+    {
+      "epoch": 0.3251852695840767,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 6377
+    },
+    {
+      "epoch": 0.32523626304018205,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 6378
+    },
+    {
+      "epoch": 0.3252872564962874,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 6379
+    },
+    {
+      "epoch": 0.32533824995239286,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32538924340849823,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 6381
+    },
+    {
+      "epoch": 0.3254402368646036,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 6382
+    },
+    {
+      "epoch": 0.325491230320709,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3255422237768144,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 6384
+    },
+    {
+      "epoch": 0.3255932172329198,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 6385
+    },
+    {
+      "epoch": 0.32564421068902516,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 6386
+    },
+    {
+      "epoch": 0.3256952041451306,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 6387
+    },
+    {
+      "epoch": 0.32574619760123597,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 6388
+    },
+    {
+      "epoch": 0.32579719105734134,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 6389
+    },
+    {
+      "epoch": 0.3258481845134467,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32589917796955215,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 6391
+    },
+    {
+      "epoch": 0.3259501714256575,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 6392
+    },
+    {
+      "epoch": 0.3260011648817629,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 6393
+    },
+    {
+      "epoch": 0.3260521583378683,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8609,
+      "step": 6394
+    },
+    {
+      "epoch": 0.3261031517939737,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 6395
+    },
+    {
+      "epoch": 0.3261541452500791,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 6396
+    },
+    {
+      "epoch": 0.32620513870618445,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 6397
+    },
+    {
+      "epoch": 0.3262561321622899,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 6398
+    },
+    {
+      "epoch": 0.32630712561839526,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 6399
+    },
+    {
+      "epoch": 0.32635811907450063,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 6400
+    },
+    {
+      "epoch": 0.32640911253060606,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 6401
+    },
+    {
+      "epoch": 0.32646010598671144,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 6402
+    },
+    {
+      "epoch": 0.3265110994428168,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 6403
+    },
+    {
+      "epoch": 0.3265620928989222,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 6404
+    },
+    {
+      "epoch": 0.3266130863550276,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 6405
+    },
+    {
+      "epoch": 0.326664079811133,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8411,
+      "step": 6406
+    },
+    {
+      "epoch": 0.32671507326723837,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 6407
+    },
+    {
+      "epoch": 0.3267660667233438,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 6408
+    },
+    {
+      "epoch": 0.32681706017944917,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 6409
+    },
+    {
+      "epoch": 0.32686805363555455,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3269190470916599,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 6411
+    },
+    {
+      "epoch": 0.32697004054776535,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 6412
+    },
+    {
+      "epoch": 0.3270210340038707,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 6413
+    },
+    {
+      "epoch": 0.3270720274599761,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 6414
+    },
+    {
+      "epoch": 0.32712302091608153,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3271740143721869,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 6416
+    },
+    {
+      "epoch": 0.3272250078282923,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 6417
+    },
+    {
+      "epoch": 0.32727600128439766,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 6418
+    },
+    {
+      "epoch": 0.3273269947405031,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 6419
+    },
+    {
+      "epoch": 0.32737798819660846,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 6420
+    },
+    {
+      "epoch": 0.32742898165271384,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 6421
+    },
+    {
+      "epoch": 0.32747997510881927,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 6422
+    },
+    {
+      "epoch": 0.32753096856492464,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 6423
+    },
+    {
+      "epoch": 0.32758196202103,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3276329554771354,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 6425
+    },
+    {
+      "epoch": 0.3276839489332408,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 6426
+    },
+    {
+      "epoch": 0.3277349423893462,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 6427
+    },
+    {
+      "epoch": 0.32778593584545157,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8517,
+      "step": 6428
+    },
+    {
+      "epoch": 0.327836929301557,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 6429
+    },
+    {
+      "epoch": 0.3278879227576624,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8433,
+      "step": 6430
+    },
+    {
+      "epoch": 0.32793891621376775,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3279899096698731,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 6432
+    },
+    {
+      "epoch": 0.32804090312597856,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 6433
+    },
+    {
+      "epoch": 0.32809189658208393,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 6434
+    },
+    {
+      "epoch": 0.3281428900381893,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 6435
+    },
+    {
+      "epoch": 0.32819388349429474,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 6436
+    },
+    {
+      "epoch": 0.3282448769504001,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 6437
+    },
+    {
+      "epoch": 0.3282958704065055,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 6438
+    },
+    {
+      "epoch": 0.32834686386261086,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 6439
+    },
+    {
+      "epoch": 0.3283978573187163,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32844885077482167,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 6441
+    },
+    {
+      "epoch": 0.32849984423092704,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 6442
+    },
+    {
+      "epoch": 0.32855083768703247,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 6443
+    },
+    {
+      "epoch": 0.32860183114313785,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 6444
+    },
+    {
+      "epoch": 0.3286528245992432,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 6445
+    },
+    {
+      "epoch": 0.3287038180553486,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 6446
+    },
+    {
+      "epoch": 0.328754811511454,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 6447
+    },
+    {
+      "epoch": 0.3288058049675594,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 6448
+    },
+    {
+      "epoch": 0.3288567984236648,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8562,
+      "step": 6449
+    },
+    {
+      "epoch": 0.3289077918797702,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3289587853358756,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 6451
+    },
+    {
+      "epoch": 0.32900977879198096,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 6452
+    },
+    {
+      "epoch": 0.32906077224808633,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 6453
+    },
+    {
+      "epoch": 0.32911176570419176,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6454
+    },
+    {
+      "epoch": 0.32916275916029714,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 6455
+    },
+    {
+      "epoch": 0.3292137526164025,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 6456
+    },
+    {
+      "epoch": 0.32926474607250794,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3293157395286133,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3293667329847187,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 6459
+    },
+    {
+      "epoch": 0.32941772644082407,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3294687198969295,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 6461
+    },
+    {
+      "epoch": 0.32951971335303487,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 6462
+    },
+    {
+      "epoch": 0.32957070680914025,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 6463
+    },
+    {
+      "epoch": 0.3296217002652457,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 6464
+    },
+    {
+      "epoch": 0.32967269372135105,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3297236871774564,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 6466
+    },
+    {
+      "epoch": 0.3297746806335618,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 6467
+    },
+    {
+      "epoch": 0.32982567408966723,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 6468
+    },
+    {
+      "epoch": 0.3298766675457726,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 6469
+    },
+    {
+      "epoch": 0.329927661001878,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3299786544579834,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 6471
+    },
+    {
+      "epoch": 0.3300296479140888,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 6472
+    },
+    {
+      "epoch": 0.33008064137019416,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 6473
+    },
+    {
+      "epoch": 0.33013163482629954,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 6474
+    },
+    {
+      "epoch": 0.33018262828240497,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 6475
+    },
+    {
+      "epoch": 0.33023362173851034,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8664,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3302846151946157,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 6477
+    },
+    {
+      "epoch": 0.33033560865072115,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 6478
+    },
+    {
+      "epoch": 0.3303866021068265,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3304375955629319,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 6480
+    },
+    {
+      "epoch": 0.33048858901903727,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 6481
+    },
+    {
+      "epoch": 0.3305395824751427,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 6482
+    },
+    {
+      "epoch": 0.3305905759312481,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6483
+    },
+    {
+      "epoch": 0.33064156938735345,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3306925628434589,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 6485
+    },
+    {
+      "epoch": 0.33074355629956426,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 6486
+    },
+    {
+      "epoch": 0.33079454975566963,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 6487
+    },
+    {
+      "epoch": 0.330845543211775,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 6488
+    },
+    {
+      "epoch": 0.33089653666788044,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3309475301239858,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3309985235800912,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3310495170361966,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 6492
+    },
+    {
+      "epoch": 0.331100510492302,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 6493
+    },
+    {
+      "epoch": 0.33115150394840737,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 6494
+    },
+    {
+      "epoch": 0.33120249740451274,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 6495
+    },
+    {
+      "epoch": 0.33125349086061817,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 6496
+    },
+    {
+      "epoch": 0.33130448431672355,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3313554777728289,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 6498
+    },
+    {
+      "epoch": 0.33140647122893435,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3314574646850397,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3315084581411451,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3315594515972505,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3316104450533559,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 6503
+    },
+    {
+      "epoch": 0.3316614385094613,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6504
+    },
+    {
+      "epoch": 0.33171243196556666,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3317634254216721,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 6506
+    },
+    {
+      "epoch": 0.33181441887777746,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 6507
+    },
+    {
+      "epoch": 0.33186541233388284,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3319164057899882,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 6509
+    },
+    {
+      "epoch": 0.33196739924609364,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 6510
+    },
+    {
+      "epoch": 0.332018392702199,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 6511
+    },
+    {
+      "epoch": 0.3320693861583044,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3321203796144098,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8437,
+      "step": 6513
+    },
+    {
+      "epoch": 0.3321713730705152,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 6514
+    },
+    {
+      "epoch": 0.33222236652662057,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 6515
+    },
+    {
+      "epoch": 0.33227335998272595,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 6516
+    },
+    {
+      "epoch": 0.3323243534388314,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 6517
+    },
+    {
+      "epoch": 0.33237534689493675,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3324263403510421,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8594,
+      "step": 6519
+    },
+    {
+      "epoch": 0.33247733380714756,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33252832726325293,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 6521
+    },
+    {
+      "epoch": 0.3325793207193583,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8636,
+      "step": 6522
+    },
+    {
+      "epoch": 0.3326303141754637,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8576,
+      "step": 6523
+    },
+    {
+      "epoch": 0.3326813076315691,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 6524
+    },
+    {
+      "epoch": 0.3327323010876745,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8612,
+      "step": 6525
+    },
+    {
+      "epoch": 0.33278329454377986,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3328342879998853,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 6527
+    },
+    {
+      "epoch": 0.33288528145599067,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 6528
+    },
+    {
+      "epoch": 0.33293627491209604,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6529
+    },
+    {
+      "epoch": 0.3329872683682014,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 6530
+    },
+    {
+      "epoch": 0.33303826182430685,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 6531
+    },
+    {
+      "epoch": 0.3330892552804122,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3331402487365176,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 6533
+    },
+    {
+      "epoch": 0.333191242192623,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3332422356487284,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3332932291048338,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 6536
+    },
+    {
+      "epoch": 0.33334422256093915,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 6537
+    },
+    {
+      "epoch": 0.3333952160170446,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 6538
+    },
+    {
+      "epoch": 0.33344620947314996,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 6539
+    },
+    {
+      "epoch": 0.33349720292925533,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 6540
+    },
+    {
+      "epoch": 0.33354819638536076,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 6541
+    },
+    {
+      "epoch": 0.33359918984146614,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 6542
+    },
+    {
+      "epoch": 0.3336501832975715,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 6543
+    },
+    {
+      "epoch": 0.3337011767536769,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 6544
+    },
+    {
+      "epoch": 0.3337521702097823,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3338031636658877,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 6546
+    },
+    {
+      "epoch": 0.33385415712199307,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3339051505780985,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3339561440342039,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 6549
+    },
+    {
+      "epoch": 0.33400713749030925,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3340581309464146,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6551
+    },
+    {
+      "epoch": 0.33410912440252005,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 6552
+    },
+    {
+      "epoch": 0.3341601178586254,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3342111113147308,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 6554
+    },
+    {
+      "epoch": 0.33426210477083623,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8669,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3343130982269416,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 6556
+    },
+    {
+      "epoch": 0.334364091683047,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 6557
+    },
+    {
+      "epoch": 0.33441508513915236,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3344660785952578,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 6559
+    },
+    {
+      "epoch": 0.33451707205136316,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8821,
+      "step": 6560
+    },
+    {
+      "epoch": 0.33456806550746854,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 6561
+    },
+    {
+      "epoch": 0.33461905896357397,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 6562
+    },
+    {
+      "epoch": 0.33467005241967934,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.861,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3347210458757847,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8688,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3347720393318901,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3348230327879955,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3348740262441009,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 6567
+    },
+    {
+      "epoch": 0.3349250197002063,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3349760131563117,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 6569
+    },
+    {
+      "epoch": 0.3350270066124171,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33507800006852245,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 6571
+    },
+    {
+      "epoch": 0.3351289935246278,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 6572
+    },
+    {
+      "epoch": 0.33517998698073326,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 6573
+    },
+    {
+      "epoch": 0.33523098043683863,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 6574
+    },
+    {
+      "epoch": 0.335281973892944,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 6575
+    },
+    {
+      "epoch": 0.33533296734904944,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8486,
+      "step": 6576
+    },
+    {
+      "epoch": 0.3353839608051548,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 6577
+    },
+    {
+      "epoch": 0.3354349542612602,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 6578
+    },
+    {
+      "epoch": 0.33548594771736556,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 6579
+    },
+    {
+      "epoch": 0.335536941173471,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33558793462957637,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 6581
+    },
+    {
+      "epoch": 0.33563892808568174,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 6582
+    },
+    {
+      "epoch": 0.3356899215417872,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 6583
+    },
+    {
+      "epoch": 0.33574091499789255,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 6584
+    },
+    {
+      "epoch": 0.3357919084539979,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3358429019101033,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 6586
+    },
+    {
+      "epoch": 0.33589389536620873,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 6587
+    },
+    {
+      "epoch": 0.3359448888223141,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7854,
+      "step": 6588
+    },
+    {
+      "epoch": 0.3359958822784195,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 6589
+    },
+    {
+      "epoch": 0.3360468757345249,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3360978691906303,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 6591
+    },
+    {
+      "epoch": 0.33614886264673566,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 6592
+    },
+    {
+      "epoch": 0.33619985610284103,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 6593
+    },
+    {
+      "epoch": 0.33625084955894646,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 6594
+    },
+    {
+      "epoch": 0.33630184301505184,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3363528364711572,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 6596
+    },
+    {
+      "epoch": 0.33640382992726264,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 6597
+    },
+    {
+      "epoch": 0.336454823383368,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 6598
+    },
+    {
+      "epoch": 0.3365058168394734,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 6599
+    },
+    {
+      "epoch": 0.33655681029557877,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3366078037516842,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 6601
+    },
+    {
+      "epoch": 0.3366587972077896,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6602
+    },
+    {
+      "epoch": 0.33670979066389495,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3367607841200004,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 6604
+    },
+    {
+      "epoch": 0.33681177757610575,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8411,
+      "step": 6605
+    },
+    {
+      "epoch": 0.33686277103221113,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 6606
+    },
+    {
+      "epoch": 0.3369137644883165,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 6607
+    },
+    {
+      "epoch": 0.33696475794442193,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 6608
+    },
+    {
+      "epoch": 0.3370157514005273,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 6609
+    },
+    {
+      "epoch": 0.3370667448566327,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8696,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3371177383127381,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 6611
+    },
+    {
+      "epoch": 0.3371687317688435,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 6612
+    },
+    {
+      "epoch": 0.33721972522494886,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 6613
+    },
+    {
+      "epoch": 0.33727071868105424,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 6614
+    },
+    {
+      "epoch": 0.33732171213715967,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6615
+    },
+    {
+      "epoch": 0.33737270559326504,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 6616
+    },
+    {
+      "epoch": 0.3374236990493704,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 6617
+    },
+    {
+      "epoch": 0.33747469250547585,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3375256859615812,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3375766794176866,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 6620
+    },
+    {
+      "epoch": 0.337627672873792,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 6621
+    },
+    {
+      "epoch": 0.3376786663298974,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 6622
+    },
+    {
+      "epoch": 0.3377296597860028,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 6623
+    },
+    {
+      "epoch": 0.33778065324210815,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 6624
+    },
+    {
+      "epoch": 0.3378316466982136,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 6625
+    },
+    {
+      "epoch": 0.33788264015431896,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 6626
+    },
+    {
+      "epoch": 0.33793363361042433,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 6627
+    },
+    {
+      "epoch": 0.3379846270665297,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 6628
+    },
+    {
+      "epoch": 0.33803562052263514,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6629
+    },
+    {
+      "epoch": 0.3380866139787405,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3381376074348459,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 6631
+    },
+    {
+      "epoch": 0.3381886008909513,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3382395943470567,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 6633
+    },
+    {
+      "epoch": 0.33829058780316207,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 6634
+    },
+    {
+      "epoch": 0.33834158125926744,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 6635
+    },
+    {
+      "epoch": 0.3383925747153729,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 6636
+    },
+    {
+      "epoch": 0.33844356817147825,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 6637
+    },
+    {
+      "epoch": 0.3384945616275836,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 6638
+    },
+    {
+      "epoch": 0.33854555508368905,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 6639
+    },
+    {
+      "epoch": 0.33859654853979443,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3386475419958998,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 6641
+    },
+    {
+      "epoch": 0.3386985354520052,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 6642
+    },
+    {
+      "epoch": 0.3387495289081106,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 6643
+    },
+    {
+      "epoch": 0.338800522364216,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 6644
+    },
+    {
+      "epoch": 0.33885151582032136,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 6645
+    },
+    {
+      "epoch": 0.3389025092764268,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 6646
+    },
+    {
+      "epoch": 0.33895350273253216,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 6647
+    },
+    {
+      "epoch": 0.33900449618863754,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 6648
+    },
+    {
+      "epoch": 0.3390554896447429,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 6649
+    },
+    {
+      "epoch": 0.33910648310084834,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8801,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3391574765569537,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 6651
+    },
+    {
+      "epoch": 0.3392084700130591,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 6652
+    },
+    {
+      "epoch": 0.33925946346916447,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 6653
+    },
+    {
+      "epoch": 0.3393104569252699,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 6654
+    },
+    {
+      "epoch": 0.3393614503813753,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 6655
+    },
+    {
+      "epoch": 0.33941244383748065,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8479,
+      "step": 6656
+    },
+    {
+      "epoch": 0.3394634372935861,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 6657
+    },
+    {
+      "epoch": 0.33951443074969145,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 6658
+    },
+    {
+      "epoch": 0.33956542420579683,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 6659
+    },
+    {
+      "epoch": 0.3396164176619022,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 6660
+    },
+    {
+      "epoch": 0.33966741111800763,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 6661
+    },
+    {
+      "epoch": 0.339718404574113,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 6662
+    },
+    {
+      "epoch": 0.3397693980302184,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 6663
+    },
+    {
+      "epoch": 0.3398203914863238,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 6664
+    },
+    {
+      "epoch": 0.3398713849424292,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 6665
+    },
+    {
+      "epoch": 0.33992237839853456,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 6666
+    },
+    {
+      "epoch": 0.33997337185463994,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6667
+    },
+    {
+      "epoch": 0.34002436531074537,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 6668
+    },
+    {
+      "epoch": 0.34007535876685074,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3401263522229561,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 6670
+    },
+    {
+      "epoch": 0.34017734567906155,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 6671
+    },
+    {
+      "epoch": 0.3402283391351669,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 6672
+    },
+    {
+      "epoch": 0.3402793325912723,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 6673
+    },
+    {
+      "epoch": 0.3403303260473777,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 6674
+    },
+    {
+      "epoch": 0.3403813195034831,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 6675
+    },
+    {
+      "epoch": 0.3404323129595885,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 6676
+    },
+    {
+      "epoch": 0.34048330641569385,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 6677
+    },
+    {
+      "epoch": 0.3405342998717993,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8607,
+      "step": 6678
+    },
+    {
+      "epoch": 0.34058529332790466,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 6679
+    },
+    {
+      "epoch": 0.34063628678401003,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3406872802401154,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 6681
+    },
+    {
+      "epoch": 0.34073827369622084,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8433,
+      "step": 6682
+    },
+    {
+      "epoch": 0.3407892671523262,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 6683
+    },
+    {
+      "epoch": 0.3408402606084316,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8515,
+      "step": 6684
+    },
+    {
+      "epoch": 0.340891254064537,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 6685
+    },
+    {
+      "epoch": 0.3409422475206424,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 6686
+    },
+    {
+      "epoch": 0.34099324097674777,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 6687
+    },
+    {
+      "epoch": 0.34104423443285314,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 6688
+    },
+    {
+      "epoch": 0.3410952278889586,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6689
+    },
+    {
+      "epoch": 0.34114622134506395,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3411972148011693,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 6691
+    },
+    {
+      "epoch": 0.34124820825727475,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 6692
+    },
+    {
+      "epoch": 0.34129920171338013,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 6693
+    },
+    {
+      "epoch": 0.3413501951694855,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 6694
+    },
+    {
+      "epoch": 0.3414011886255909,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 6695
+    },
+    {
+      "epoch": 0.3414521820816963,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 6696
+    },
+    {
+      "epoch": 0.3415031755378017,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 6697
+    },
+    {
+      "epoch": 0.34155416899390706,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 6698
+    },
+    {
+      "epoch": 0.3416051624500125,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 6699
+    },
+    {
+      "epoch": 0.34165615590611786,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34170714936222324,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8566,
+      "step": 6701
+    },
+    {
+      "epoch": 0.3417581428183286,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 6702
+    },
+    {
+      "epoch": 0.34180913627443404,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 6703
+    },
+    {
+      "epoch": 0.3418601297305394,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6704
+    },
+    {
+      "epoch": 0.3419111231866448,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 6705
+    },
+    {
+      "epoch": 0.3419621166427502,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 6706
+    },
+    {
+      "epoch": 0.3420131100988556,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8662,
+      "step": 6707
+    },
+    {
+      "epoch": 0.342064103554961,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 6708
+    },
+    {
+      "epoch": 0.34211509701106635,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 6709
+    },
+    {
+      "epoch": 0.3421660904671718,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34221708392327715,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 6711
+    },
+    {
+      "epoch": 0.34226807737938253,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 6712
+    },
+    {
+      "epoch": 0.34231907083548796,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 6713
+    },
+    {
+      "epoch": 0.34237006429159333,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8572,
+      "step": 6714
+    },
+    {
+      "epoch": 0.3424210577476987,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 6715
+    },
+    {
+      "epoch": 0.3424720512038041,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 6716
+    },
+    {
+      "epoch": 0.3425230446599095,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 6717
+    },
+    {
+      "epoch": 0.3425740381160149,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8692,
+      "step": 6718
+    },
+    {
+      "epoch": 0.34262503157212026,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6719
+    },
+    {
+      "epoch": 0.3426760250282257,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34272701848433107,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 6721
+    },
+    {
+      "epoch": 0.34277801194043644,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 6722
+    },
+    {
+      "epoch": 0.3428290053965418,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 6723
+    },
+    {
+      "epoch": 0.34287999885264725,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 6724
+    },
+    {
+      "epoch": 0.3429309923087526,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 6725
+    },
+    {
+      "epoch": 0.342981985764858,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 6726
+    },
+    {
+      "epoch": 0.34303297922096343,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 6727
+    },
+    {
+      "epoch": 0.3430839726770688,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 6728
+    },
+    {
+      "epoch": 0.3431349661331742,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 6729
+    },
+    {
+      "epoch": 0.34318595958927955,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8641,
+      "step": 6730
+    },
+    {
+      "epoch": 0.343236953045385,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 6731
+    },
+    {
+      "epoch": 0.34328794650149036,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6732
+    },
+    {
+      "epoch": 0.34333893995759573,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6733
+    },
+    {
+      "epoch": 0.34338993341370116,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 6734
+    },
+    {
+      "epoch": 0.34344092686980654,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6735
+    },
+    {
+      "epoch": 0.3434919203259119,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 6736
+    },
+    {
+      "epoch": 0.3435429137820173,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 6737
+    },
+    {
+      "epoch": 0.3435939072381227,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 6738
+    },
+    {
+      "epoch": 0.3436449006942281,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 6739
+    },
+    {
+      "epoch": 0.34369589415033347,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3437468876064389,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 6741
+    },
+    {
+      "epoch": 0.3437978810625443,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 6742
+    },
+    {
+      "epoch": 0.34384887451864965,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 6743
+    },
+    {
+      "epoch": 0.343899867974755,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 6744
+    },
+    {
+      "epoch": 0.34395086143086046,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 6745
+    },
+    {
+      "epoch": 0.34400185488696583,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 6746
+    },
+    {
+      "epoch": 0.3440528483430712,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 6747
+    },
+    {
+      "epoch": 0.34410384179917664,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8791,
+      "step": 6748
+    },
+    {
+      "epoch": 0.344154835255282,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8586,
+      "step": 6749
+    },
+    {
+      "epoch": 0.3442058287113874,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 6750
+    },
+    {
+      "epoch": 0.34425682216749276,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 6751
+    },
+    {
+      "epoch": 0.3443078156235982,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 6752
+    },
+    {
+      "epoch": 0.34435880907970357,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 6753
+    },
+    {
+      "epoch": 0.34440980253580894,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 6754
+    },
+    {
+      "epoch": 0.34446079599191437,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 6755
+    },
+    {
+      "epoch": 0.34451178944801975,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 6756
+    },
+    {
+      "epoch": 0.3445627829041251,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 6757
+    },
+    {
+      "epoch": 0.3446137763602305,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 6758
+    },
+    {
+      "epoch": 0.3446647698163359,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 6759
+    },
+    {
+      "epoch": 0.3447157632724413,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3447667567285467,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 6761
+    },
+    {
+      "epoch": 0.3448177501846521,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 6762
+    },
+    {
+      "epoch": 0.3448687436407575,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 6763
+    },
+    {
+      "epoch": 0.34491973709686286,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 6764
+    },
+    {
+      "epoch": 0.34497073055296823,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 6765
+    },
+    {
+      "epoch": 0.34502172400907366,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 6766
+    },
+    {
+      "epoch": 0.34507271746517904,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 6767
+    },
+    {
+      "epoch": 0.3451237109212844,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 6768
+    },
+    {
+      "epoch": 0.34517470437738984,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 6769
+    },
+    {
+      "epoch": 0.3452256978334952,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3452766912896006,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 6771
+    },
+    {
+      "epoch": 0.34532768474570597,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 6772
+    },
+    {
+      "epoch": 0.3453786782018114,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 6773
+    },
+    {
+      "epoch": 0.34542967165791677,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 6774
+    },
+    {
+      "epoch": 0.34548066511402215,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3455316585701276,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 6776
+    },
+    {
+      "epoch": 0.34558265202623295,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 6777
+    },
+    {
+      "epoch": 0.3456336454823383,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 6778
+    },
+    {
+      "epoch": 0.3456846389384437,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 6779
+    },
+    {
+      "epoch": 0.34573563239454913,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3457866258506545,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 6781
+    },
+    {
+      "epoch": 0.3458376193067599,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6782
+    },
+    {
+      "epoch": 0.3458886127628653,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 6783
+    },
+    {
+      "epoch": 0.3459396062189707,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 6784
+    },
+    {
+      "epoch": 0.34599059967507606,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 6785
+    },
+    {
+      "epoch": 0.34604159313118144,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 6786
+    },
+    {
+      "epoch": 0.34609258658728687,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 6787
+    },
+    {
+      "epoch": 0.34614358004339224,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 6788
+    },
+    {
+      "epoch": 0.3461945734994976,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 6789
+    },
+    {
+      "epoch": 0.34624556695560305,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3462965604117084,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 6791
+    },
+    {
+      "epoch": 0.3463475538678138,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 6792
+    },
+    {
+      "epoch": 0.34639854732391917,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8535,
+      "step": 6793
+    },
+    {
+      "epoch": 0.3464495407800246,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 6794
+    },
+    {
+      "epoch": 0.34650053423613,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 6795
+    },
+    {
+      "epoch": 0.34655152769223535,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 6796
+    },
+    {
+      "epoch": 0.3466025211483408,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 6797
+    },
+    {
+      "epoch": 0.34665351460444616,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 6798
+    },
+    {
+      "epoch": 0.34670450806055153,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 6799
+    },
+    {
+      "epoch": 0.3467555015166569,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 6800
+    },
+    {
+      "epoch": 0.34680649497276234,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 6801
+    },
+    {
+      "epoch": 0.3468574884288677,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 6802
+    },
+    {
+      "epoch": 0.3469084818849731,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 6803
+    },
+    {
+      "epoch": 0.3469594753410785,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 6804
+    },
+    {
+      "epoch": 0.3470104687971839,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 6805
+    },
+    {
+      "epoch": 0.34706146225328927,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 6806
+    },
+    {
+      "epoch": 0.34711245570939464,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 6807
+    },
+    {
+      "epoch": 0.34716344916550007,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 6808
+    },
+    {
+      "epoch": 0.34721444262160545,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 6809
+    },
+    {
+      "epoch": 0.3472654360777108,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 6810
+    },
+    {
+      "epoch": 0.34731642953381625,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 6811
+    },
+    {
+      "epoch": 0.3473674229899216,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 6812
+    },
+    {
+      "epoch": 0.347418416446027,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 6813
+    },
+    {
+      "epoch": 0.3474694099021324,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 6814
+    },
+    {
+      "epoch": 0.3475204033582378,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 6815
+    },
+    {
+      "epoch": 0.3475713968143432,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 6816
+    },
+    {
+      "epoch": 0.34762239027044856,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6817
+    },
+    {
+      "epoch": 0.347673383726554,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6818
+    },
+    {
+      "epoch": 0.34772437718265936,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 6819
+    },
+    {
+      "epoch": 0.34777537063876474,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3478263640948701,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 6821
+    },
+    {
+      "epoch": 0.34787735755097554,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 6822
+    },
+    {
+      "epoch": 0.3479283510070809,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 6823
+    },
+    {
+      "epoch": 0.3479793444631863,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 6824
+    },
+    {
+      "epoch": 0.3480303379192917,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 6825
+    },
+    {
+      "epoch": 0.3480813313753971,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 6826
+    },
+    {
+      "epoch": 0.34813232483150247,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 6827
+    },
+    {
+      "epoch": 0.34818331828760785,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3482343117437133,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 6829
+    },
+    {
+      "epoch": 0.34828530519981865,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 6830
+    },
+    {
+      "epoch": 0.348336298655924,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 6831
+    },
+    {
+      "epoch": 0.34838729211202946,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 6832
+    },
+    {
+      "epoch": 0.34843828556813483,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 6833
+    },
+    {
+      "epoch": 0.3484892790242402,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 6834
+    },
+    {
+      "epoch": 0.3485402724803456,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 6835
+    },
+    {
+      "epoch": 0.348591265936451,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 6836
+    },
+    {
+      "epoch": 0.3486422593925564,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 6837
+    },
+    {
+      "epoch": 0.34869325284866176,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8708,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3487442463047672,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 6839
+    },
+    {
+      "epoch": 0.34879523976087257,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 6840
+    },
+    {
+      "epoch": 0.34884623321697794,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 6841
+    },
+    {
+      "epoch": 0.3488972266730833,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 6842
+    },
+    {
+      "epoch": 0.34894822012918875,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 6843
+    },
+    {
+      "epoch": 0.3489992135852941,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 6844
+    },
+    {
+      "epoch": 0.3490502070413995,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 6845
+    },
+    {
+      "epoch": 0.3491012004975049,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3491521939536103,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 6847
+    },
+    {
+      "epoch": 0.3492031874097157,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8552,
+      "step": 6848
+    },
+    {
+      "epoch": 0.34925418086582105,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 6849
+    },
+    {
+      "epoch": 0.3493051743219265,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34935616777803186,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6851
+    },
+    {
+      "epoch": 0.34940716123413723,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 6852
+    },
+    {
+      "epoch": 0.34945815469024266,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6853
+    },
+    {
+      "epoch": 0.34950914814634804,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6854
+    },
+    {
+      "epoch": 0.3495601416024534,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 6855
+    },
+    {
+      "epoch": 0.3496111350585588,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 6856
+    },
+    {
+      "epoch": 0.3496621285146642,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 6857
+    },
+    {
+      "epoch": 0.3497131219707696,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8663,
+      "step": 6858
+    },
+    {
+      "epoch": 0.34976411542687497,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 6859
+    },
+    {
+      "epoch": 0.3498151088829804,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 6860
+    },
+    {
+      "epoch": 0.34986610233908577,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 6861
+    },
+    {
+      "epoch": 0.34991709579519115,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3499680892512965,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 6863
+    },
+    {
+      "epoch": 0.35001908270740195,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 6864
+    },
+    {
+      "epoch": 0.3500700761635073,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8537,
+      "step": 6865
+    },
+    {
+      "epoch": 0.3501210696196127,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 6866
+    },
+    {
+      "epoch": 0.35017206307571813,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 6867
+    },
+    {
+      "epoch": 0.3502230565318235,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 6868
+    },
+    {
+      "epoch": 0.3502740499879289,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 6869
+    },
+    {
+      "epoch": 0.35032504344403426,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3503760369001397,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8571,
+      "step": 6871
+    },
+    {
+      "epoch": 0.35042703035624506,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 6872
+    },
+    {
+      "epoch": 0.35047802381235044,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6873
+    },
+    {
+      "epoch": 0.35052901726845587,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 6874
+    },
+    {
+      "epoch": 0.35058001072456124,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3506310041806666,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 6876
+    },
+    {
+      "epoch": 0.350681997636772,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 6877
+    },
+    {
+      "epoch": 0.3507329910928774,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3507839845489828,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6879
+    },
+    {
+      "epoch": 0.35083497800508817,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3508859714611936,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 6881
+    },
+    {
+      "epoch": 0.350936964917299,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 6882
+    },
+    {
+      "epoch": 0.35098795837340435,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3510389518295097,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 6884
+    },
+    {
+      "epoch": 0.35108994528561516,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 6885
+    },
+    {
+      "epoch": 0.35114093874172053,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3511919321978259,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 6887
+    },
+    {
+      "epoch": 0.35124292565393134,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 6888
+    },
+    {
+      "epoch": 0.3512939191100367,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 6889
+    },
+    {
+      "epoch": 0.3513449125661421,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 6890
+    },
+    {
+      "epoch": 0.35139590602224746,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 6891
+    },
+    {
+      "epoch": 0.3514468994783529,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6892
+    },
+    {
+      "epoch": 0.35149789293445827,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 6893
+    },
+    {
+      "epoch": 0.35154888639056364,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 6894
+    },
+    {
+      "epoch": 0.35159987984666907,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 6895
+    },
+    {
+      "epoch": 0.35165087330277445,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 6896
+    },
+    {
+      "epoch": 0.3517018667588798,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 6897
+    },
+    {
+      "epoch": 0.3517528602149852,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 6898
+    },
+    {
+      "epoch": 0.3518038536710906,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6899
+    },
+    {
+      "epoch": 0.351854847127196,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3519058405833014,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6901
+    },
+    {
+      "epoch": 0.3519568340394068,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3520078274955122,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 6903
+    },
+    {
+      "epoch": 0.35205882095161756,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 6904
+    },
+    {
+      "epoch": 0.35210981440772293,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 6905
+    },
+    {
+      "epoch": 0.35216080786382836,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 6906
+    },
+    {
+      "epoch": 0.35221180131993374,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8812,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3522627947760391,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 6908
+    },
+    {
+      "epoch": 0.35231378823214454,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 6909
+    },
+    {
+      "epoch": 0.3523647816882499,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3524157751443553,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 6911
+    },
+    {
+      "epoch": 0.35246676860046067,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3525177620565661,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3525687555126715,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 6914
+    },
+    {
+      "epoch": 0.35261974896877685,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 6915
+    },
+    {
+      "epoch": 0.3526707424248823,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 6916
+    },
+    {
+      "epoch": 0.35272173588098765,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 6917
+    },
+    {
+      "epoch": 0.352772729337093,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8659,
+      "step": 6918
+    },
+    {
+      "epoch": 0.3528237227931984,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 6919
+    },
+    {
+      "epoch": 0.35287471624930383,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3529257097054092,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 6921
+    },
+    {
+      "epoch": 0.3529767031615146,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 6922
+    },
+    {
+      "epoch": 0.35302769661762,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 6923
+    },
+    {
+      "epoch": 0.3530786900737254,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 6924
+    },
+    {
+      "epoch": 0.35312968352983076,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 6925
+    },
+    {
+      "epoch": 0.35318067698593614,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 6926
+    },
+    {
+      "epoch": 0.35323167044204157,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 6927
+    },
+    {
+      "epoch": 0.35328266389814694,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3533336573542523,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 6929
+    },
+    {
+      "epoch": 0.35338465081035775,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3534356442664631,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3534866377225685,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 6932
+    },
+    {
+      "epoch": 0.3535376311786739,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 6933
+    },
+    {
+      "epoch": 0.3535886246347793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3536396180908847,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8621,
+      "step": 6935
+    },
+    {
+      "epoch": 0.35369061154699005,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3537416050030955,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 6937
+    },
+    {
+      "epoch": 0.35379259845920086,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 6938
+    },
+    {
+      "epoch": 0.35384359191530623,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 6939
+    },
+    {
+      "epoch": 0.3538945853714116,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 6940
+    },
+    {
+      "epoch": 0.35394557882751704,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3539965722836224,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 6942
+    },
+    {
+      "epoch": 0.3540475657397278,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 6943
+    },
+    {
+      "epoch": 0.3540985591958332,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 6944
+    },
+    {
+      "epoch": 0.3541495526519386,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 6945
+    },
+    {
+      "epoch": 0.35420054610804397,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 6946
+    },
+    {
+      "epoch": 0.35425153956414934,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 6947
+    },
+    {
+      "epoch": 0.3543025330202548,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 6948
+    },
+    {
+      "epoch": 0.35435352647636015,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8578,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3544045199324655,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 6950
+    },
+    {
+      "epoch": 0.35445551338857095,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 6951
+    },
+    {
+      "epoch": 0.35450650684467633,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3545575003007817,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 6953
+    },
+    {
+      "epoch": 0.3546084937568871,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 6954
+    },
+    {
+      "epoch": 0.3546594872129925,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 6955
+    },
+    {
+      "epoch": 0.3547104806690979,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 6956
+    },
+    {
+      "epoch": 0.35476147412520326,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 6957
+    },
+    {
+      "epoch": 0.3548124675813087,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 6958
+    },
+    {
+      "epoch": 0.35486346103741406,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 6959
+    },
+    {
+      "epoch": 0.35491445449351944,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8384,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3549654479496248,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 6961
+    },
+    {
+      "epoch": 0.35501644140573024,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8532,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3550674348618356,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 6963
+    },
+    {
+      "epoch": 0.355118428317941,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3551694217740464,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3552204152301518,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 6966
+    },
+    {
+      "epoch": 0.3552714086862572,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 6967
+    },
+    {
+      "epoch": 0.35532240214236255,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 6968
+    },
+    {
+      "epoch": 0.355373395598468,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 6969
+    },
+    {
+      "epoch": 0.35542438905457335,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35547538251067873,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 6971
+    },
+    {
+      "epoch": 0.35552637596678416,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 6972
+    },
+    {
+      "epoch": 0.35557736942288953,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 6973
+    },
+    {
+      "epoch": 0.3556283628789949,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 6974
+    },
+    {
+      "epoch": 0.3556793563351003,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 6975
+    },
+    {
+      "epoch": 0.3557303497912057,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 6976
+    },
+    {
+      "epoch": 0.3557813432473111,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 6977
+    },
+    {
+      "epoch": 0.35583233670341646,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3558833301595219,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 6979
+    },
+    {
+      "epoch": 0.35593432361562727,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35598531707173264,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 6981
+    },
+    {
+      "epoch": 0.356036310527838,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 6982
+    },
+    {
+      "epoch": 0.35608730398394345,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 6983
+    },
+    {
+      "epoch": 0.3561382974400488,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 6984
+    },
+    {
+      "epoch": 0.3561892908961542,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 6985
+    },
+    {
+      "epoch": 0.35624028435225963,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 6986
+    },
+    {
+      "epoch": 0.356291277808365,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 6987
+    },
+    {
+      "epoch": 0.3563422712644704,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 6988
+    },
+    {
+      "epoch": 0.35639326472057575,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 6989
+    },
+    {
+      "epoch": 0.3564442581766812,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 6990
+    },
+    {
+      "epoch": 0.35649525163278656,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 6991
+    },
+    {
+      "epoch": 0.35654624508889193,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 6992
+    },
+    {
+      "epoch": 0.35659723854499736,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 6993
+    },
+    {
+      "epoch": 0.35664823200110274,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3566992254572081,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 6995
+    },
+    {
+      "epoch": 0.3567502189133135,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 6996
+    },
+    {
+      "epoch": 0.3568012123694189,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 6997
+    },
+    {
+      "epoch": 0.3568522058255243,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 6998
+    },
+    {
+      "epoch": 0.35690319928162967,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 6999
+    },
+    {
+      "epoch": 0.3569541927377351,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8539,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3570051861938405,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 7001
+    },
+    {
+      "epoch": 0.35705617964994585,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 7002
+    },
+    {
+      "epoch": 0.3571071731060512,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 7003
+    },
+    {
+      "epoch": 0.35715816656215665,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 7004
+    },
+    {
+      "epoch": 0.35720916001826203,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3572601534743674,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 7006
+    },
+    {
+      "epoch": 0.35731114693047283,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8611,
+      "step": 7007
+    },
+    {
+      "epoch": 0.3573621403865782,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.859,
+      "step": 7008
+    },
+    {
+      "epoch": 0.3574131338426836,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 7009
+    },
+    {
+      "epoch": 0.35746412729878896,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 7010
+    },
+    {
+      "epoch": 0.3575151207548944,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 7011
+    },
+    {
+      "epoch": 0.35756611421099976,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 7012
+    },
+    {
+      "epoch": 0.35761710766710514,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 7013
+    },
+    {
+      "epoch": 0.35766810112321057,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 7014
+    },
+    {
+      "epoch": 0.35771909457931594,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3577700880354213,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 7016
+    },
+    {
+      "epoch": 0.3578210814915267,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 7017
+    },
+    {
+      "epoch": 0.3578720749476321,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8484,
+      "step": 7018
+    },
+    {
+      "epoch": 0.3579230684037375,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 7019
+    },
+    {
+      "epoch": 0.3579740618598429,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3580250553159483,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 7021
+    },
+    {
+      "epoch": 0.3580760487720537,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 7022
+    },
+    {
+      "epoch": 0.35812704222815905,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 7023
+    },
+    {
+      "epoch": 0.35817803568426443,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 7024
+    },
+    {
+      "epoch": 0.35822902914036986,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 7025
+    },
+    {
+      "epoch": 0.35828002259647523,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 7026
+    },
+    {
+      "epoch": 0.3583310160525806,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 7027
+    },
+    {
+      "epoch": 0.35838200950868604,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 7028
+    },
+    {
+      "epoch": 0.3584330029647914,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7029
+    },
+    {
+      "epoch": 0.3584839964208968,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 7030
+    },
+    {
+      "epoch": 0.35853498987700216,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 7031
+    },
+    {
+      "epoch": 0.3585859833331076,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 7032
+    },
+    {
+      "epoch": 0.35863697678921297,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 7033
+    },
+    {
+      "epoch": 0.35868797024531834,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 7034
+    },
+    {
+      "epoch": 0.3587389637014238,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 7035
+    },
+    {
+      "epoch": 0.35878995715752915,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 7036
+    },
+    {
+      "epoch": 0.3588409506136345,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 7037
+    },
+    {
+      "epoch": 0.3588919440697399,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 7038
+    },
+    {
+      "epoch": 0.35894293752584533,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 7039
+    },
+    {
+      "epoch": 0.3589939309819507,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3590449244380561,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 7041
+    },
+    {
+      "epoch": 0.3590959178941615,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 7042
+    },
+    {
+      "epoch": 0.3591469113502669,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 7043
+    },
+    {
+      "epoch": 0.35919790480637226,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 7044
+    },
+    {
+      "epoch": 0.35924889826247763,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7045
+    },
+    {
+      "epoch": 0.35929989171858306,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 7046
+    },
+    {
+      "epoch": 0.35935088517468844,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 7047
+    },
+    {
+      "epoch": 0.3594018786307938,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 7048
+    },
+    {
+      "epoch": 0.35945287208689924,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7049
+    },
+    {
+      "epoch": 0.3595038655430046,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 7050
+    },
+    {
+      "epoch": 0.35955485899911,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 7051
+    },
+    {
+      "epoch": 0.35960585245521537,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 7052
+    },
+    {
+      "epoch": 0.3596568459113208,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 7053
+    },
+    {
+      "epoch": 0.3597078393674262,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 7054
+    },
+    {
+      "epoch": 0.35975883282353155,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8511,
+      "step": 7055
+    },
+    {
+      "epoch": 0.359809826279637,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 7056
+    },
+    {
+      "epoch": 0.35986081973574235,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 7057
+    },
+    {
+      "epoch": 0.35991181319184773,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 7058
+    },
+    {
+      "epoch": 0.3599628066479531,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 7059
+    },
+    {
+      "epoch": 0.36001380010405853,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3600647935601639,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 7061
+    },
+    {
+      "epoch": 0.3601157870162693,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 7062
+    },
+    {
+      "epoch": 0.3601667804723747,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 7063
+    },
+    {
+      "epoch": 0.3602177739284801,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 7064
+    },
+    {
+      "epoch": 0.36026876738458546,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 7065
+    },
+    {
+      "epoch": 0.36031976084069084,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 7066
+    },
+    {
+      "epoch": 0.36037075429679627,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 7067
+    },
+    {
+      "epoch": 0.36042174775290164,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 7068
+    },
+    {
+      "epoch": 0.360472741209007,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 7069
+    },
+    {
+      "epoch": 0.36052373466511245,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3605747281212178,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 7071
+    },
+    {
+      "epoch": 0.3606257215773232,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 7072
+    },
+    {
+      "epoch": 0.3606767150334286,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 7073
+    },
+    {
+      "epoch": 0.360727708489534,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 7074
+    },
+    {
+      "epoch": 0.3607787019456394,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 7075
+    },
+    {
+      "epoch": 0.36082969540174475,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 7076
+    },
+    {
+      "epoch": 0.3608806888578502,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 7077
+    },
+    {
+      "epoch": 0.36093168231395556,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 7078
+    },
+    {
+      "epoch": 0.36098267577006093,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7079
+    },
+    {
+      "epoch": 0.3610336692261663,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 7080
+    },
+    {
+      "epoch": 0.36108466268227174,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 7081
+    },
+    {
+      "epoch": 0.3611356561383771,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 7082
+    },
+    {
+      "epoch": 0.3611866495944825,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 7083
+    },
+    {
+      "epoch": 0.3612376430505879,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 7084
+    },
+    {
+      "epoch": 0.3612886365066933,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 7085
+    },
+    {
+      "epoch": 0.36133962996279867,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 7086
+    },
+    {
+      "epoch": 0.36139062341890404,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 7087
+    },
+    {
+      "epoch": 0.3614416168750095,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 7088
+    },
+    {
+      "epoch": 0.36149261033111485,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 7089
+    },
+    {
+      "epoch": 0.3615436037872202,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36159459724332566,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 7091
+    },
+    {
+      "epoch": 0.36164559069943103,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 7092
+    },
+    {
+      "epoch": 0.3616965841555364,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 7093
+    },
+    {
+      "epoch": 0.3617475776116418,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 7094
+    },
+    {
+      "epoch": 0.3617985710677472,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3618495645238526,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 7096
+    },
+    {
+      "epoch": 0.36190055797995796,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7097
+    },
+    {
+      "epoch": 0.3619515514360634,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 7098
+    },
+    {
+      "epoch": 0.36200254489216876,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 7099
+    },
+    {
+      "epoch": 0.36205353834827414,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3621045318043795,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 7101
+    },
+    {
+      "epoch": 0.36215552526048495,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.843,
+      "step": 7102
+    },
+    {
+      "epoch": 0.3622065187165903,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 7103
+    },
+    {
+      "epoch": 0.3622575121726957,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 7104
+    },
+    {
+      "epoch": 0.3623085056288011,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 7105
+    },
+    {
+      "epoch": 0.3623594990849065,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 7106
+    },
+    {
+      "epoch": 0.3624104925410119,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 7107
+    },
+    {
+      "epoch": 0.36246148599711725,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 7108
+    },
+    {
+      "epoch": 0.3625124794532227,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8711,
+      "step": 7109
+    },
+    {
+      "epoch": 0.36256347290932806,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36261446636543343,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 7111
+    },
+    {
+      "epoch": 0.36266545982153886,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 7112
+    },
+    {
+      "epoch": 0.36271645327764424,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7113
+    },
+    {
+      "epoch": 0.3627674467337496,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 7114
+    },
+    {
+      "epoch": 0.362818440189855,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 7115
+    },
+    {
+      "epoch": 0.3628694336459604,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 7116
+    },
+    {
+      "epoch": 0.3629204271020658,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 7117
+    },
+    {
+      "epoch": 0.36297142055817117,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 7118
+    },
+    {
+      "epoch": 0.3630224140142766,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 7119
+    },
+    {
+      "epoch": 0.36307340747038197,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 7120
+    },
+    {
+      "epoch": 0.36312440092648735,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 7121
+    },
+    {
+      "epoch": 0.3631753943825927,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 7122
+    },
+    {
+      "epoch": 0.36322638783869815,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 7123
+    },
+    {
+      "epoch": 0.3632773812948035,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3633283747509089,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 7125
+    },
+    {
+      "epoch": 0.36337936820701433,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3634303616631197,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 7127
+    },
+    {
+      "epoch": 0.3634813551192251,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 7128
+    },
+    {
+      "epoch": 0.36353234857533046,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 7129
+    },
+    {
+      "epoch": 0.3635833420314359,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 7130
+    },
+    {
+      "epoch": 0.36363433548754126,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 7131
+    },
+    {
+      "epoch": 0.36368532894364664,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 7132
+    },
+    {
+      "epoch": 0.36373632239975207,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 7133
+    },
+    {
+      "epoch": 0.36378731585585744,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3638383093119628,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3638893027680682,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3639402962241736,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 7137
+    },
+    {
+      "epoch": 0.363991289680279,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 7138
+    },
+    {
+      "epoch": 0.36404228313638437,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3640932765924898,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3641442700485952,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 7141
+    },
+    {
+      "epoch": 0.36419526350470055,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 7142
+    },
+    {
+      "epoch": 0.3642462569608059,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 7143
+    },
+    {
+      "epoch": 0.36429725041691136,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 7144
+    },
+    {
+      "epoch": 0.36434824387301673,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3643992373291221,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8498,
+      "step": 7146
+    },
+    {
+      "epoch": 0.36445023078522754,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 7147
+    },
+    {
+      "epoch": 0.3645012242413329,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7148
+    },
+    {
+      "epoch": 0.3645522176974383,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 7149
+    },
+    {
+      "epoch": 0.36460321115354366,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3646542046096491,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7151
+    },
+    {
+      "epoch": 0.36470519806575447,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 7152
+    },
+    {
+      "epoch": 0.36475619152185984,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 7153
+    },
+    {
+      "epoch": 0.36480718497796527,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 7154
+    },
+    {
+      "epoch": 0.36485817843407065,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8489,
+      "step": 7155
+    },
+    {
+      "epoch": 0.364909171890176,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8596,
+      "step": 7156
+    },
+    {
+      "epoch": 0.3649601653462814,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 7157
+    },
+    {
+      "epoch": 0.3650111588023868,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 7158
+    },
+    {
+      "epoch": 0.3650621522584922,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 7159
+    },
+    {
+      "epoch": 0.3651131457145976,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 7160
+    },
+    {
+      "epoch": 0.365164139170703,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 7161
+    },
+    {
+      "epoch": 0.3652151326268084,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 7162
+    },
+    {
+      "epoch": 0.36526612608291376,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 7163
+    },
+    {
+      "epoch": 0.36531711953901913,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 7164
+    },
+    {
+      "epoch": 0.36536811299512456,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 7165
+    },
+    {
+      "epoch": 0.36541910645122994,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3654700999073353,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7167
+    },
+    {
+      "epoch": 0.36552109336344074,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3655720868195461,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.853,
+      "step": 7169
+    },
+    {
+      "epoch": 0.3656230802756515,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36567407373175687,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 7171
+    },
+    {
+      "epoch": 0.3657250671878623,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 7172
+    },
+    {
+      "epoch": 0.36577606064396767,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 7173
+    },
+    {
+      "epoch": 0.36582705410007305,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3658780475561785,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 7175
+    },
+    {
+      "epoch": 0.36592904101228385,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3659800344683892,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7177
+    },
+    {
+      "epoch": 0.3660310279244946,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7178
+    },
+    {
+      "epoch": 0.36608202138060003,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 7179
+    },
+    {
+      "epoch": 0.3661330148367054,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3661840082928108,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3662350017489162,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 7182
+    },
+    {
+      "epoch": 0.3662859952050216,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 7183
+    },
+    {
+      "epoch": 0.36633698866112696,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 7184
+    },
+    {
+      "epoch": 0.36638798211723234,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 7185
+    },
+    {
+      "epoch": 0.36643897557333777,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 7186
+    },
+    {
+      "epoch": 0.36648996902944314,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 7187
+    },
+    {
+      "epoch": 0.3665409624855485,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 7188
+    },
+    {
+      "epoch": 0.36659195594165395,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 7189
+    },
+    {
+      "epoch": 0.3666429493977593,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3666939428538647,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 7191
+    },
+    {
+      "epoch": 0.36674493630997007,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3667959297660755,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 7193
+    },
+    {
+      "epoch": 0.3668469232221809,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 7194
+    },
+    {
+      "epoch": 0.36689791667828625,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 7195
+    },
+    {
+      "epoch": 0.3669489101343917,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 7196
+    },
+    {
+      "epoch": 0.36699990359049706,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 7197
+    },
+    {
+      "epoch": 0.36705089704660243,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 7198
+    },
+    {
+      "epoch": 0.3671018905027078,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 7199
+    },
+    {
+      "epoch": 0.36715288395881324,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3672038774149186,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 7201
+    },
+    {
+      "epoch": 0.367254870871024,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 7202
+    },
+    {
+      "epoch": 0.3673058643271294,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 7203
+    },
+    {
+      "epoch": 0.3673568577832348,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 7204
+    },
+    {
+      "epoch": 0.36740785123934017,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 7205
+    },
+    {
+      "epoch": 0.36745884469544554,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 7206
+    },
+    {
+      "epoch": 0.36750983815155097,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 7207
+    },
+    {
+      "epoch": 0.36756083160765635,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 7208
+    },
+    {
+      "epoch": 0.3676118250637617,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 7209
+    },
+    {
+      "epoch": 0.36766281851986715,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3677138119759725,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 7211
+    },
+    {
+      "epoch": 0.3677648054320779,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 7212
+    },
+    {
+      "epoch": 0.3678157988881833,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3678667923442887,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 7214
+    },
+    {
+      "epoch": 0.3679177858003941,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 7215
+    },
+    {
+      "epoch": 0.36796877925649946,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3680197727126049,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 7217
+    },
+    {
+      "epoch": 0.36807076616871026,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 7218
+    },
+    {
+      "epoch": 0.36812175962481564,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 7219
+    },
+    {
+      "epoch": 0.368172753080921,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 7220
+    },
+    {
+      "epoch": 0.36822374653702644,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3682747399931318,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 7222
+    },
+    {
+      "epoch": 0.3683257334492372,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3683767269053426,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 7224
+    },
+    {
+      "epoch": 0.368427720361448,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 7225
+    },
+    {
+      "epoch": 0.36847871381755337,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 7226
+    },
+    {
+      "epoch": 0.36852970727365875,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 7227
+    },
+    {
+      "epoch": 0.3685807007297642,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 7228
+    },
+    {
+      "epoch": 0.36863169418586955,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8617,
+      "step": 7229
+    },
+    {
+      "epoch": 0.3686826876419749,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 7230
+    },
+    {
+      "epoch": 0.36873368109808036,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 7231
+    },
+    {
+      "epoch": 0.36878467455418573,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8546,
+      "step": 7232
+    },
+    {
+      "epoch": 0.3688356680102911,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 7233
+    },
+    {
+      "epoch": 0.3688866614663965,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 7234
+    },
+    {
+      "epoch": 0.3689376549225019,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 7235
+    },
+    {
+      "epoch": 0.3689886483786073,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 7236
+    },
+    {
+      "epoch": 0.36903964183471266,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 7237
+    },
+    {
+      "epoch": 0.3690906352908181,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 7238
+    },
+    {
+      "epoch": 0.36914162874692347,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 7239
+    },
+    {
+      "epoch": 0.36919262220302884,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3692436156591342,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 7241
+    },
+    {
+      "epoch": 0.36929460911523965,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 7242
+    },
+    {
+      "epoch": 0.369345602571345,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 7243
+    },
+    {
+      "epoch": 0.3693965960274504,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 7244
+    },
+    {
+      "epoch": 0.3694475894835558,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 7245
+    },
+    {
+      "epoch": 0.3694985829396612,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 7246
+    },
+    {
+      "epoch": 0.3695495763957666,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 7247
+    },
+    {
+      "epoch": 0.36960056985187195,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 7248
+    },
+    {
+      "epoch": 0.3696515633079774,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 7249
+    },
+    {
+      "epoch": 0.36970255676408276,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 7250
+    },
+    {
+      "epoch": 0.36975355022018813,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 7251
+    },
+    {
+      "epoch": 0.36980454367629356,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 7252
+    },
+    {
+      "epoch": 0.36985553713239894,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 7253
+    },
+    {
+      "epoch": 0.3699065305885043,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 7254
+    },
+    {
+      "epoch": 0.3699575240446097,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3700085175007151,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3700595109568205,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7257
+    },
+    {
+      "epoch": 0.37011050441292587,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3701614978690313,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 7259
+    },
+    {
+      "epoch": 0.37021249132513667,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 7260
+    },
+    {
+      "epoch": 0.37026348478124205,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 7261
+    },
+    {
+      "epoch": 0.3703144782373474,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 7262
+    },
+    {
+      "epoch": 0.37036547169345285,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 7263
+    },
+    {
+      "epoch": 0.3704164651495582,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 7264
+    },
+    {
+      "epoch": 0.3704674586056636,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 7265
+    },
+    {
+      "epoch": 0.37051845206176903,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 7266
+    },
+    {
+      "epoch": 0.3705694455178744,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8457,
+      "step": 7267
+    },
+    {
+      "epoch": 0.3706204389739798,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 7268
+    },
+    {
+      "epoch": 0.37067143243008516,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 7269
+    },
+    {
+      "epoch": 0.3707224258861906,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 7270
+    },
+    {
+      "epoch": 0.37077341934229596,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 7271
+    },
+    {
+      "epoch": 0.37082441279840134,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 7272
+    },
+    {
+      "epoch": 0.37087540625450677,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 7273
+    },
+    {
+      "epoch": 0.37092639971061214,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3709773931667175,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 7275
+    },
+    {
+      "epoch": 0.3710283866228229,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 7276
+    },
+    {
+      "epoch": 0.3710793800789283,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 7277
+    },
+    {
+      "epoch": 0.3711303735350337,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 7278
+    },
+    {
+      "epoch": 0.3711813669911391,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 7279
+    },
+    {
+      "epoch": 0.3712323604472445,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3712833539033499,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 7281
+    },
+    {
+      "epoch": 0.37133434735945525,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8584,
+      "step": 7282
+    },
+    {
+      "epoch": 0.3713853408155606,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 7283
+    },
+    {
+      "epoch": 0.37143633427166606,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 7284
+    },
+    {
+      "epoch": 0.37148732772777143,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 7285
+    },
+    {
+      "epoch": 0.3715383211838768,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 7286
+    },
+    {
+      "epoch": 0.37158931463998224,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 7287
+    },
+    {
+      "epoch": 0.3716403080960876,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 7288
+    },
+    {
+      "epoch": 0.371691301552193,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 7289
+    },
+    {
+      "epoch": 0.37174229500829836,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3717932884644038,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 7291
+    },
+    {
+      "epoch": 0.37184428192050917,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 7292
+    },
+    {
+      "epoch": 0.37189527537661454,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 7293
+    },
+    {
+      "epoch": 0.37194626883272,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 7294
+    },
+    {
+      "epoch": 0.37199726228882535,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3720482557449307,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 7296
+    },
+    {
+      "epoch": 0.3720992492010361,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8409,
+      "step": 7297
+    },
+    {
+      "epoch": 0.37215024265714153,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 7298
+    },
+    {
+      "epoch": 0.3722012361132469,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 7299
+    },
+    {
+      "epoch": 0.3722522295693523,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3723032230254577,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 7301
+    },
+    {
+      "epoch": 0.3723542164815631,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 7302
+    },
+    {
+      "epoch": 0.37240520993766846,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 7303
+    },
+    {
+      "epoch": 0.37245620339377383,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 7304
+    },
+    {
+      "epoch": 0.37250719684987926,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 7305
+    },
+    {
+      "epoch": 0.37255819030598464,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 7306
+    },
+    {
+      "epoch": 0.37260918376209,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 7307
+    },
+    {
+      "epoch": 0.37266017721819544,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 7308
+    },
+    {
+      "epoch": 0.3727111706743008,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7309
+    },
+    {
+      "epoch": 0.3727621641304062,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 7310
+    },
+    {
+      "epoch": 0.37281315758651157,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 7311
+    },
+    {
+      "epoch": 0.372864151042617,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 7312
+    },
+    {
+      "epoch": 0.3729151444987224,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 7313
+    },
+    {
+      "epoch": 0.37296613795482775,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7314
+    },
+    {
+      "epoch": 0.3730171314109332,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8516,
+      "step": 7315
+    },
+    {
+      "epoch": 0.37306812486703855,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 7316
+    },
+    {
+      "epoch": 0.37311911832314393,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 7317
+    },
+    {
+      "epoch": 0.3731701117792493,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 7318
+    },
+    {
+      "epoch": 0.37322110523535473,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 7319
+    },
+    {
+      "epoch": 0.3732720986914601,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3733230921475655,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 7321
+    },
+    {
+      "epoch": 0.3733740856036709,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 7322
+    },
+    {
+      "epoch": 0.3734250790597763,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 7323
+    },
+    {
+      "epoch": 0.37347607251588166,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 7324
+    },
+    {
+      "epoch": 0.37352706597198704,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 7325
+    },
+    {
+      "epoch": 0.37357805942809247,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 7326
+    },
+    {
+      "epoch": 0.37362905288419784,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.848,
+      "step": 7327
+    },
+    {
+      "epoch": 0.3736800463403032,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7328
+    },
+    {
+      "epoch": 0.37373103979640865,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8443,
+      "step": 7329
+    },
+    {
+      "epoch": 0.373782033252514,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3738330267086194,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7331
+    },
+    {
+      "epoch": 0.3738840201647248,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 7332
+    },
+    {
+      "epoch": 0.3739350136208302,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3739860070769356,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 7334
+    },
+    {
+      "epoch": 0.37403700053304095,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 7335
+    },
+    {
+      "epoch": 0.3740879939891464,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 7336
+    },
+    {
+      "epoch": 0.37413898744525176,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 7337
+    },
+    {
+      "epoch": 0.37418998090135713,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 7338
+    },
+    {
+      "epoch": 0.3742409743574625,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 7339
+    },
+    {
+      "epoch": 0.37429196781356794,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 7340
+    },
+    {
+      "epoch": 0.3743429612696733,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 7341
+    },
+    {
+      "epoch": 0.3743939547257787,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 7342
+    },
+    {
+      "epoch": 0.3744449481818841,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 7343
+    },
+    {
+      "epoch": 0.3744959416379895,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 7344
+    },
+    {
+      "epoch": 0.37454693509409487,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 7345
+    },
+    {
+      "epoch": 0.37459792855020024,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 7346
+    },
+    {
+      "epoch": 0.3746489220063057,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 7347
+    },
+    {
+      "epoch": 0.37469991546241105,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 7348
+    },
+    {
+      "epoch": 0.3747509089185164,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 7349
+    },
+    {
+      "epoch": 0.37480190237462185,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37485289583072723,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 7351
+    },
+    {
+      "epoch": 0.3749038892868326,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 7352
+    },
+    {
+      "epoch": 0.374954882742938,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 7353
+    },
+    {
+      "epoch": 0.3750058761990434,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 7354
+    },
+    {
+      "epoch": 0.3750568696551488,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 7355
+    },
+    {
+      "epoch": 0.37510786311125416,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3751588565673596,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 7357
+    },
+    {
+      "epoch": 0.37520985002346496,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 7358
+    },
+    {
+      "epoch": 0.37526084347957034,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 7359
+    },
+    {
+      "epoch": 0.3753118369356757,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 7360
+    },
+    {
+      "epoch": 0.37536283039178114,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 7361
+    },
+    {
+      "epoch": 0.3754138238478865,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8501,
+      "step": 7362
+    },
+    {
+      "epoch": 0.3754648173039919,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 7363
+    },
+    {
+      "epoch": 0.3755158107600973,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 7364
+    },
+    {
+      "epoch": 0.3755668042162027,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8543,
+      "step": 7365
+    },
+    {
+      "epoch": 0.3756177976723081,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 7366
+    },
+    {
+      "epoch": 0.37566879112841345,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 7367
+    },
+    {
+      "epoch": 0.3757197845845189,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 7368
+    },
+    {
+      "epoch": 0.37577077804062425,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 7369
+    },
+    {
+      "epoch": 0.37582177149672963,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 7370
+    },
+    {
+      "epoch": 0.37587276495283506,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 7371
+    },
+    {
+      "epoch": 0.37592375840894043,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 7372
+    },
+    {
+      "epoch": 0.3759747518650458,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 7373
+    },
+    {
+      "epoch": 0.3760257453211512,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 7374
+    },
+    {
+      "epoch": 0.3760767387772566,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 7375
+    },
+    {
+      "epoch": 0.376127732233362,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 7376
+    },
+    {
+      "epoch": 0.37617872568946736,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3762297191455728,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 7378
+    },
+    {
+      "epoch": 0.37628071260167817,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 7379
+    },
+    {
+      "epoch": 0.37633170605778354,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3763826995138889,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 7381
+    },
+    {
+      "epoch": 0.37643369296999435,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 7382
+    },
+    {
+      "epoch": 0.3764846864260997,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 7383
+    },
+    {
+      "epoch": 0.3765356798822051,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8534,
+      "step": 7384
+    },
+    {
+      "epoch": 0.37658667333831053,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3766376667944159,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 7386
+    },
+    {
+      "epoch": 0.3766886602505213,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7387
+    },
+    {
+      "epoch": 0.37673965370662665,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 7388
+    },
+    {
+      "epoch": 0.3767906471627321,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 7389
+    },
+    {
+      "epoch": 0.37684164061883746,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7390
+    },
+    {
+      "epoch": 0.37689263407494283,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 7391
+    },
+    {
+      "epoch": 0.37694362753104826,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 7392
+    },
+    {
+      "epoch": 0.37699462098715364,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 7393
+    },
+    {
+      "epoch": 0.377045614443259,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 7394
+    },
+    {
+      "epoch": 0.3770966078993644,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 7395
+    },
+    {
+      "epoch": 0.3771476013554698,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 7396
+    },
+    {
+      "epoch": 0.3771985948115752,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 7397
+    },
+    {
+      "epoch": 0.37724958826768057,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 7398
+    },
+    {
+      "epoch": 0.377300581723786,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 7399
+    },
+    {
+      "epoch": 0.3773515751798914,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8467,
+      "step": 7400
+    },
+    {
+      "epoch": 0.37740256863599675,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7401
+    },
+    {
+      "epoch": 0.3774535620921021,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 7402
+    },
+    {
+      "epoch": 0.37750455554820755,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 7403
+    },
+    {
+      "epoch": 0.37755554900431293,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3776065424604183,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 7405
+    },
+    {
+      "epoch": 0.37765753591652373,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 7406
+    },
+    {
+      "epoch": 0.3777085293726291,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 7407
+    },
+    {
+      "epoch": 0.3777595228287345,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 7408
+    },
+    {
+      "epoch": 0.37781051628483986,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 7409
+    },
+    {
+      "epoch": 0.3778615097409453,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 7410
+    },
+    {
+      "epoch": 0.37791250319705066,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 7411
+    },
+    {
+      "epoch": 0.37796349665315604,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 7412
+    },
+    {
+      "epoch": 0.37801449010926147,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 7413
+    },
+    {
+      "epoch": 0.37806548356536684,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 7414
+    },
+    {
+      "epoch": 0.3781164770214722,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3781674704775776,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 7416
+    },
+    {
+      "epoch": 0.378218463933683,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8509,
+      "step": 7417
+    },
+    {
+      "epoch": 0.3782694573897884,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 7418
+    },
+    {
+      "epoch": 0.3783204508458938,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 7419
+    },
+    {
+      "epoch": 0.3783714443019992,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3784224377581046,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 7421
+    },
+    {
+      "epoch": 0.37847343121420995,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 7422
+    },
+    {
+      "epoch": 0.37852442467031533,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 7423
+    },
+    {
+      "epoch": 0.37857541812642076,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 7424
+    },
+    {
+      "epoch": 0.37862641158252613,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3786774050386315,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 7426
+    },
+    {
+      "epoch": 0.37872839849473694,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 7427
+    },
+    {
+      "epoch": 0.3787793919508423,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 7428
+    },
+    {
+      "epoch": 0.3788303854069477,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 7429
+    },
+    {
+      "epoch": 0.37888137886305306,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3789323723191585,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 7431
+    },
+    {
+      "epoch": 0.37898336577526387,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 7432
+    },
+    {
+      "epoch": 0.37903435923136924,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.85,
+      "step": 7433
+    },
+    {
+      "epoch": 0.3790853526874747,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 7434
+    },
+    {
+      "epoch": 0.37913634614358005,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 7435
+    },
+    {
+      "epoch": 0.3791873395996854,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7436
+    },
+    {
+      "epoch": 0.3792383330557908,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 7437
+    },
+    {
+      "epoch": 0.37928932651189623,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 7438
+    },
+    {
+      "epoch": 0.3793403199680016,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 7439
+    },
+    {
+      "epoch": 0.379391313424107,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3794423068802124,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 7441
+    },
+    {
+      "epoch": 0.3794933003363178,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 7442
+    },
+    {
+      "epoch": 0.37954429379242316,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 7443
+    },
+    {
+      "epoch": 0.37959528724852853,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 7444
+    },
+    {
+      "epoch": 0.37964628070463396,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 7445
+    },
+    {
+      "epoch": 0.37969727416073934,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 7446
+    },
+    {
+      "epoch": 0.3797482676168447,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 7447
+    },
+    {
+      "epoch": 0.37979926107295015,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3798502545290555,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 7449
+    },
+    {
+      "epoch": 0.3799012479851609,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 7450
+    },
+    {
+      "epoch": 0.37995224144126627,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 7451
+    },
+    {
+      "epoch": 0.3800032348973717,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 7452
+    },
+    {
+      "epoch": 0.3800542283534771,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 7453
+    },
+    {
+      "epoch": 0.38010522180958245,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8403,
+      "step": 7454
+    },
+    {
+      "epoch": 0.3801562152656879,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 7455
+    },
+    {
+      "epoch": 0.38020720872179326,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 7456
+    },
+    {
+      "epoch": 0.38025820217789863,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7457
+    },
+    {
+      "epoch": 0.380309195634004,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 7458
+    },
+    {
+      "epoch": 0.38036018909010944,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 7459
+    },
+    {
+      "epoch": 0.3804111825462148,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3804621760023202,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 7461
+    },
+    {
+      "epoch": 0.3805131694584256,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 7462
+    },
+    {
+      "epoch": 0.380564162914531,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 7463
+    },
+    {
+      "epoch": 0.38061515637063636,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 7464
+    },
+    {
+      "epoch": 0.38066614982674174,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 7465
+    },
+    {
+      "epoch": 0.38071714328284717,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 7466
+    },
+    {
+      "epoch": 0.38076813673895255,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 7467
+    },
+    {
+      "epoch": 0.3808191301950579,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 7468
+    },
+    {
+      "epoch": 0.38087012365116335,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 7469
+    },
+    {
+      "epoch": 0.3809211171072687,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3809721105633741,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7471
+    },
+    {
+      "epoch": 0.3810231040194795,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 7472
+    },
+    {
+      "epoch": 0.3810740974755849,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 7473
+    },
+    {
+      "epoch": 0.3811250909316903,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 7474
+    },
+    {
+      "epoch": 0.38117608438779566,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3812270778439011,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 7476
+    },
+    {
+      "epoch": 0.38127807130000646,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 7477
+    },
+    {
+      "epoch": 0.38132906475611184,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 7478
+    },
+    {
+      "epoch": 0.3813800582122172,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 7479
+    },
+    {
+      "epoch": 0.38143105166832264,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 7480
+    },
+    {
+      "epoch": 0.381482045124428,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 7481
+    },
+    {
+      "epoch": 0.3815330385805334,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 7482
+    },
+    {
+      "epoch": 0.3815840320366388,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8499,
+      "step": 7483
+    },
+    {
+      "epoch": 0.3816350254927442,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 7484
+    },
+    {
+      "epoch": 0.38168601894884957,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 7485
+    },
+    {
+      "epoch": 0.38173701240495495,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 7486
+    },
+    {
+      "epoch": 0.3817880058610604,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 7487
+    },
+    {
+      "epoch": 0.38183899931716575,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3818899927732711,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 7489
+    },
+    {
+      "epoch": 0.38194098622937656,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 7490
+    },
+    {
+      "epoch": 0.38199197968548193,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 7491
+    },
+    {
+      "epoch": 0.3820429731415873,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 7492
+    },
+    {
+      "epoch": 0.3820939665976927,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 7493
+    },
+    {
+      "epoch": 0.3821449600537981,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 7494
+    },
+    {
+      "epoch": 0.3821959535099035,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 7495
+    },
+    {
+      "epoch": 0.38224694696600886,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7624,
+      "step": 7496
+    },
+    {
+      "epoch": 0.3822979404221143,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 7497
+    },
+    {
+      "epoch": 0.38234893387821967,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 7498
+    },
+    {
+      "epoch": 0.38239992733432504,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 7499
+    },
+    {
+      "epoch": 0.3824509207904304,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 7500
+    },
+    {
+      "epoch": 0.38250191424653585,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 7501
+    },
+    {
+      "epoch": 0.3825529077026412,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 7502
+    },
+    {
+      "epoch": 0.3826039011587466,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 7503
+    },
+    {
+      "epoch": 0.382654894614852,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8567,
+      "step": 7504
+    },
+    {
+      "epoch": 0.3827058880709574,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3827568815270628,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 7506
+    },
+    {
+      "epoch": 0.38280787498316815,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 7507
+    },
+    {
+      "epoch": 0.3828588684392736,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 7508
+    },
+    {
+      "epoch": 0.38290986189537896,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8545,
+      "step": 7509
+    },
+    {
+      "epoch": 0.38296085535148433,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 7510
+    },
+    {
+      "epoch": 0.38301184880758976,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 7511
+    },
+    {
+      "epoch": 0.38306284226369514,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 7512
+    },
+    {
+      "epoch": 0.3831138357198005,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 7513
+    },
+    {
+      "epoch": 0.3831648291759059,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 7514
+    },
+    {
+      "epoch": 0.3832158226320113,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 7515
+    },
+    {
+      "epoch": 0.3832668160881167,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 7516
+    },
+    {
+      "epoch": 0.38331780954422207,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 7517
+    },
+    {
+      "epoch": 0.3833688030003275,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 7518
+    },
+    {
+      "epoch": 0.38341979645643287,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 7519
+    },
+    {
+      "epoch": 0.38347078991253825,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3835217833686436,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 7521
+    },
+    {
+      "epoch": 0.38357277682474905,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3836237702808544,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3836747637369598,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 7524
+    },
+    {
+      "epoch": 0.38372575719306523,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3837767506491706,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 7526
+    },
+    {
+      "epoch": 0.383827744105276,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 7527
+    },
+    {
+      "epoch": 0.38387873756138136,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 7528
+    },
+    {
+      "epoch": 0.3839297310174868,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7529
+    },
+    {
+      "epoch": 0.38398072447359216,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38403171792969754,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 7531
+    },
+    {
+      "epoch": 0.38408271138580297,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 7532
+    },
+    {
+      "epoch": 0.38413370484190834,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 7533
+    },
+    {
+      "epoch": 0.3841846982980137,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 7534
+    },
+    {
+      "epoch": 0.3842356917541191,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 7535
+    },
+    {
+      "epoch": 0.3842866852102245,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7536
+    },
+    {
+      "epoch": 0.3843376786663299,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 7537
+    },
+    {
+      "epoch": 0.38438867212243527,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3844396655785407,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 7539
+    },
+    {
+      "epoch": 0.3844906590346461,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38454165249075145,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 7541
+    },
+    {
+      "epoch": 0.3845926459468568,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 7542
+    },
+    {
+      "epoch": 0.38464363940296226,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 7543
+    },
+    {
+      "epoch": 0.38469463285906763,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 7544
+    },
+    {
+      "epoch": 0.384745626315173,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 7545
+    },
+    {
+      "epoch": 0.38479661977127844,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 7546
+    },
+    {
+      "epoch": 0.3848476132273838,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3848986066834892,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 7548
+    },
+    {
+      "epoch": 0.38494960013959456,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 7549
+    },
+    {
+      "epoch": 0.3850005935957,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 7550
+    },
+    {
+      "epoch": 0.38505158705180537,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 7551
+    },
+    {
+      "epoch": 0.38510258050791074,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 7552
+    },
+    {
+      "epoch": 0.38515357396401617,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 7553
+    },
+    {
+      "epoch": 0.38520456742012155,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 7554
+    },
+    {
+      "epoch": 0.3852555608762269,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3853065543323323,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 7556
+    },
+    {
+      "epoch": 0.3853575477884377,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 7557
+    },
+    {
+      "epoch": 0.3854085412445431,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 7558
+    },
+    {
+      "epoch": 0.3854595347006485,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 7559
+    },
+    {
+      "epoch": 0.3855105281567539,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3855615216128593,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 7561
+    },
+    {
+      "epoch": 0.38561251506896466,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 7562
+    },
+    {
+      "epoch": 0.38566350852507003,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 7563
+    },
+    {
+      "epoch": 0.38571450198117546,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 7564
+    },
+    {
+      "epoch": 0.38576549543728084,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3858164888933862,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 7566
+    },
+    {
+      "epoch": 0.38586748234949164,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 7567
+    },
+    {
+      "epoch": 0.385918475805597,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 7568
+    },
+    {
+      "epoch": 0.3859694692617024,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 7569
+    },
+    {
+      "epoch": 0.38602046271780777,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3860714561739132,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8445,
+      "step": 7571
+    },
+    {
+      "epoch": 0.38612244963001857,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 7572
+    },
+    {
+      "epoch": 0.38617344308612395,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 7573
+    },
+    {
+      "epoch": 0.3862244365422294,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 7574
+    },
+    {
+      "epoch": 0.38627542999833475,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3863264234544401,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 7576
+    },
+    {
+      "epoch": 0.3863774169105455,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 7577
+    },
+    {
+      "epoch": 0.38642841036665093,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 7578
+    },
+    {
+      "epoch": 0.3864794038227563,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 7579
+    },
+    {
+      "epoch": 0.3865303972788617,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3865813907349671,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 7581
+    },
+    {
+      "epoch": 0.3866323841910725,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 7582
+    },
+    {
+      "epoch": 0.38668337764717786,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 7583
+    },
+    {
+      "epoch": 0.38673437110328324,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 7584
+    },
+    {
+      "epoch": 0.38678536455938867,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 7585
+    },
+    {
+      "epoch": 0.38683635801549404,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 7586
+    },
+    {
+      "epoch": 0.3868873514715994,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 7587
+    },
+    {
+      "epoch": 0.38693834492770485,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 7588
+    },
+    {
+      "epoch": 0.3869893383838102,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 7589
+    },
+    {
+      "epoch": 0.3870403318399156,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 7590
+    },
+    {
+      "epoch": 0.38709132529602097,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 7591
+    },
+    {
+      "epoch": 0.3871423187521264,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8503,
+      "step": 7592
+    },
+    {
+      "epoch": 0.3871933122082318,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 7593
+    },
+    {
+      "epoch": 0.38724430566433715,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 7594
+    },
+    {
+      "epoch": 0.3872952991204426,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 7595
+    },
+    {
+      "epoch": 0.38734629257654796,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 7596
+    },
+    {
+      "epoch": 0.38739728603265333,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8661,
+      "step": 7597
+    },
+    {
+      "epoch": 0.3874482794887587,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 7598
+    },
+    {
+      "epoch": 0.38749927294486414,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 7599
+    },
+    {
+      "epoch": 0.3875502664009695,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3876012598570749,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 7601
+    },
+    {
+      "epoch": 0.3876522533131803,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 7602
+    },
+    {
+      "epoch": 0.3877032467692857,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 7603
+    },
+    {
+      "epoch": 0.38775424022539107,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 7604
+    },
+    {
+      "epoch": 0.38780523368149644,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 7605
+    },
+    {
+      "epoch": 0.38785622713760187,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 7606
+    },
+    {
+      "epoch": 0.38790722059370725,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 7607
+    },
+    {
+      "epoch": 0.3879582140498126,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7608
+    },
+    {
+      "epoch": 0.38800920750591805,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 7609
+    },
+    {
+      "epoch": 0.3880602009620234,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3881111944181288,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 7611
+    },
+    {
+      "epoch": 0.3881621878742342,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3882131813303396,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8488,
+      "step": 7613
+    },
+    {
+      "epoch": 0.388264174786445,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 7614
+    },
+    {
+      "epoch": 0.38831516824255036,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3883661616986558,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 7616
+    },
+    {
+      "epoch": 0.38841715515476116,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 7617
+    },
+    {
+      "epoch": 0.38846814861086654,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 7618
+    },
+    {
+      "epoch": 0.3885191420669719,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 7619
+    },
+    {
+      "epoch": 0.38857013552307734,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3886211289791827,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 7621
+    },
+    {
+      "epoch": 0.3886721224352881,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 7622
+    },
+    {
+      "epoch": 0.3887231158913935,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 7623
+    },
+    {
+      "epoch": 0.3887741093474989,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 7624
+    },
+    {
+      "epoch": 0.38882510280360427,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 7625
+    },
+    {
+      "epoch": 0.38887609625970965,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 7626
+    },
+    {
+      "epoch": 0.3889270897158151,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7627
+    },
+    {
+      "epoch": 0.38897808317192045,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7628
+    },
+    {
+      "epoch": 0.3890290766280258,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 7629
+    },
+    {
+      "epoch": 0.38908007008413126,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 7630
+    },
+    {
+      "epoch": 0.38913106354023663,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 7631
+    },
+    {
+      "epoch": 0.389182056996342,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 7632
+    },
+    {
+      "epoch": 0.3892330504524474,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 7633
+    },
+    {
+      "epoch": 0.3892840439085528,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 7634
+    },
+    {
+      "epoch": 0.3893350373646582,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 7635
+    },
+    {
+      "epoch": 0.38938603082076356,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 7636
+    },
+    {
+      "epoch": 0.389437024276869,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 7637
+    },
+    {
+      "epoch": 0.38948801773297437,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8533,
+      "step": 7638
+    },
+    {
+      "epoch": 0.38953901118907974,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7639
+    },
+    {
+      "epoch": 0.3895900046451851,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 7640
+    },
+    {
+      "epoch": 0.38964099810129055,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 7641
+    },
+    {
+      "epoch": 0.3896919915573959,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 7642
+    },
+    {
+      "epoch": 0.3897429850135013,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 7643
+    },
+    {
+      "epoch": 0.38979397846960673,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 7644
+    },
+    {
+      "epoch": 0.3898449719257121,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 7645
+    },
+    {
+      "epoch": 0.3898959653818175,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 7646
+    },
+    {
+      "epoch": 0.38994695883792285,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 7647
+    },
+    {
+      "epoch": 0.3899979522940283,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 7648
+    },
+    {
+      "epoch": 0.39004894575013366,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 7649
+    },
+    {
+      "epoch": 0.39009993920623903,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 7650
+    },
+    {
+      "epoch": 0.39015093266234446,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7651
+    },
+    {
+      "epoch": 0.39020192611844984,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 7652
+    },
+    {
+      "epoch": 0.3902529195745552,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 7653
+    },
+    {
+      "epoch": 0.3903039130306606,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 7654
+    },
+    {
+      "epoch": 0.390354906486766,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 7655
+    },
+    {
+      "epoch": 0.3904058999428714,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 7656
+    },
+    {
+      "epoch": 0.39045689339897677,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3905078868550822,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8568,
+      "step": 7658
+    },
+    {
+      "epoch": 0.3905588803111876,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8633,
+      "step": 7659
+    },
+    {
+      "epoch": 0.39060987376729295,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3906608672233983,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 7661
+    },
+    {
+      "epoch": 0.39071186067950375,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 7662
+    },
+    {
+      "epoch": 0.39076285413560913,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 7663
+    },
+    {
+      "epoch": 0.3908138475917145,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 7664
+    },
+    {
+      "epoch": 0.39086484104781993,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3909158345039253,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 7666
+    },
+    {
+      "epoch": 0.3909668279600307,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 7667
+    },
+    {
+      "epoch": 0.39101782141613606,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 7668
+    },
+    {
+      "epoch": 0.3910688148722415,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 7669
+    },
+    {
+      "epoch": 0.39111980832834686,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39117080178445224,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 7671
+    },
+    {
+      "epoch": 0.39122179524055767,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 7672
+    },
+    {
+      "epoch": 0.39127278869666304,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 7673
+    },
+    {
+      "epoch": 0.3913237821527684,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 7674
+    },
+    {
+      "epoch": 0.3913747756088738,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3914257690649792,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 7676
+    },
+    {
+      "epoch": 0.3914767625210846,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 7677
+    },
+    {
+      "epoch": 0.39152775597719,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 7678
+    },
+    {
+      "epoch": 0.3915787494332954,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 7679
+    },
+    {
+      "epoch": 0.3916297428894008,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39168073634550615,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 7681
+    },
+    {
+      "epoch": 0.39173172980161153,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 7682
+    },
+    {
+      "epoch": 0.39178272325771696,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 7683
+    },
+    {
+      "epoch": 0.39183371671382233,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3918847101699277,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8469,
+      "step": 7685
+    },
+    {
+      "epoch": 0.39193570362603314,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3919866970821385,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3920376905382439,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 7688
+    },
+    {
+      "epoch": 0.39208868399434926,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 7689
+    },
+    {
+      "epoch": 0.3921396774504547,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 7690
+    },
+    {
+      "epoch": 0.39219067090656007,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 7691
+    },
+    {
+      "epoch": 0.39224166436266544,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 7692
+    },
+    {
+      "epoch": 0.3922926578187709,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 7693
+    },
+    {
+      "epoch": 0.39234365127487625,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 7694
+    },
+    {
+      "epoch": 0.3923946447309816,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 7695
+    },
+    {
+      "epoch": 0.392445638187087,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 7696
+    },
+    {
+      "epoch": 0.39249663164319243,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 7697
+    },
+    {
+      "epoch": 0.3925476250992978,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8492,
+      "step": 7698
+    },
+    {
+      "epoch": 0.3925986185554032,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 7699
+    },
+    {
+      "epoch": 0.3926496120115086,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8519,
+      "step": 7700
+    },
+    {
+      "epoch": 0.392700605467614,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 7701
+    },
+    {
+      "epoch": 0.39275159892371936,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 7702
+    },
+    {
+      "epoch": 0.39280259237982473,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 7703
+    },
+    {
+      "epoch": 0.39285358583593016,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 7704
+    },
+    {
+      "epoch": 0.39290457929203554,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 7705
+    },
+    {
+      "epoch": 0.3929555727481409,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 7706
+    },
+    {
+      "epoch": 0.39300656620424634,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 7707
+    },
+    {
+      "epoch": 0.3930575596603517,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 7708
+    },
+    {
+      "epoch": 0.3931085531164571,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 7709
+    },
+    {
+      "epoch": 0.39315954657256247,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3932105400286679,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 7711
+    },
+    {
+      "epoch": 0.3932615334847733,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 7712
+    },
+    {
+      "epoch": 0.39331252694087865,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 7713
+    },
+    {
+      "epoch": 0.3933635203969841,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 7714
+    },
+    {
+      "epoch": 0.39341451385308945,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 7715
+    },
+    {
+      "epoch": 0.39346550730919483,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 7716
+    },
+    {
+      "epoch": 0.3935165007653002,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 7717
+    },
+    {
+      "epoch": 0.39356749422140563,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 7718
+    },
+    {
+      "epoch": 0.393618487677511,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 7719
+    },
+    {
+      "epoch": 0.3936694811336164,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3937204745897218,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 7721
+    },
+    {
+      "epoch": 0.3937714680458272,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 7722
+    },
+    {
+      "epoch": 0.39382246150193256,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 7723
+    },
+    {
+      "epoch": 0.39387345495803794,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 7724
+    },
+    {
+      "epoch": 0.39392444841414337,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 7725
+    },
+    {
+      "epoch": 0.39397544187024874,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 7726
+    },
+    {
+      "epoch": 0.3940264353263541,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 7727
+    },
+    {
+      "epoch": 0.39407742878245955,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 7728
+    },
+    {
+      "epoch": 0.3941284222385649,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 7729
+    },
+    {
+      "epoch": 0.3941794156946703,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3942304091507757,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 7731
+    },
+    {
+      "epoch": 0.3942814026068811,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3943323960629865,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 7733
+    },
+    {
+      "epoch": 0.39438338951909185,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 7734
+    },
+    {
+      "epoch": 0.3944343829751973,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 7735
+    },
+    {
+      "epoch": 0.39448537643130266,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 7736
+    },
+    {
+      "epoch": 0.39453636988740803,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 7737
+    },
+    {
+      "epoch": 0.3945873633435134,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 7738
+    },
+    {
+      "epoch": 0.39463835679961884,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 7739
+    },
+    {
+      "epoch": 0.3946893502557242,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3947403437118296,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 7741
+    },
+    {
+      "epoch": 0.394791337167935,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 7742
+    },
+    {
+      "epoch": 0.3948423306240404,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 7743
+    },
+    {
+      "epoch": 0.39489332408014577,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 7744
+    },
+    {
+      "epoch": 0.39494431753625114,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 7745
+    },
+    {
+      "epoch": 0.3949953109923566,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 7746
+    },
+    {
+      "epoch": 0.39504630444846195,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 7747
+    },
+    {
+      "epoch": 0.3950972979045673,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8618,
+      "step": 7748
+    },
+    {
+      "epoch": 0.39514829136067275,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 7749
+    },
+    {
+      "epoch": 0.39519928481677813,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3952502782728835,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7784,
+      "step": 7751
+    },
+    {
+      "epoch": 0.3953012717289889,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3953522651850943,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 7753
+    },
+    {
+      "epoch": 0.3954032586411997,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 7754
+    },
+    {
+      "epoch": 0.39545425209730506,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 7755
+    },
+    {
+      "epoch": 0.3955052455534105,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 7756
+    },
+    {
+      "epoch": 0.39555623900951586,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 7757
+    },
+    {
+      "epoch": 0.39560723246562124,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 7758
+    },
+    {
+      "epoch": 0.3956582259217266,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 7759
+    },
+    {
+      "epoch": 0.39570921937783204,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3957602128339374,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 7761
+    },
+    {
+      "epoch": 0.3958112062900428,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 7762
+    },
+    {
+      "epoch": 0.3958621997461482,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 7763
+    },
+    {
+      "epoch": 0.3959131932022536,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 7764
+    },
+    {
+      "epoch": 0.395964186658359,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 7765
+    },
+    {
+      "epoch": 0.39601518011446435,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 7766
+    },
+    {
+      "epoch": 0.3960661735705698,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 7767
+    },
+    {
+      "epoch": 0.39611716702667515,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 7768
+    },
+    {
+      "epoch": 0.39616816048278053,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 7769
+    },
+    {
+      "epoch": 0.39621915393888596,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 7770
+    },
+    {
+      "epoch": 0.39627014739499133,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 7771
+    },
+    {
+      "epoch": 0.3963211408510967,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3963721343072021,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 7773
+    },
+    {
+      "epoch": 0.3964231277633075,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3964741212194129,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 7775
+    },
+    {
+      "epoch": 0.39652511467551826,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3965761081316237,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 7777
+    },
+    {
+      "epoch": 0.39662710158772907,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 7778
+    },
+    {
+      "epoch": 0.39667809504383444,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 7779
+    },
+    {
+      "epoch": 0.3967290884999398,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 7780
+    },
+    {
+      "epoch": 0.39678008195604525,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 7781
+    },
+    {
+      "epoch": 0.3968310754121506,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 7782
+    },
+    {
+      "epoch": 0.396882068868256,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 7783
+    },
+    {
+      "epoch": 0.39693306232436143,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 7784
+    },
+    {
+      "epoch": 0.3969840557804668,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 7785
+    },
+    {
+      "epoch": 0.3970350492365722,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 7786
+    },
+    {
+      "epoch": 0.39708604269267755,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 7787
+    },
+    {
+      "epoch": 0.397137036148783,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 7788
+    },
+    {
+      "epoch": 0.39718802960488836,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 7789
+    },
+    {
+      "epoch": 0.39723902306099373,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 7790
+    },
+    {
+      "epoch": 0.39729001651709916,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 7791
+    },
+    {
+      "epoch": 0.39734100997320454,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 7792
+    },
+    {
+      "epoch": 0.3973920034293099,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 7793
+    },
+    {
+      "epoch": 0.3974429968854153,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 7794
+    },
+    {
+      "epoch": 0.3974939903415207,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 7795
+    },
+    {
+      "epoch": 0.3975449837976261,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 7796
+    },
+    {
+      "epoch": 0.39759597725373147,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 7797
+    },
+    {
+      "epoch": 0.3976469707098369,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 7798
+    },
+    {
+      "epoch": 0.3976979641659423,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8377,
+      "step": 7799
+    },
+    {
+      "epoch": 0.39774895762204765,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8573,
+      "step": 7800
+    },
+    {
+      "epoch": 0.397799951078153,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 7801
+    },
+    {
+      "epoch": 0.39785094453425845,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 7802
+    },
+    {
+      "epoch": 0.39790193799036383,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 7803
+    },
+    {
+      "epoch": 0.3979529314464692,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 7804
+    },
+    {
+      "epoch": 0.39800392490257464,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 7805
+    },
+    {
+      "epoch": 0.39805491835868,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 7806
+    },
+    {
+      "epoch": 0.3981059118147854,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 7807
+    },
+    {
+      "epoch": 0.39815690527089076,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 7808
+    },
+    {
+      "epoch": 0.3982078987269962,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 7809
+    },
+    {
+      "epoch": 0.39825889218310156,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 7810
+    },
+    {
+      "epoch": 0.39830988563920694,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 7811
+    },
+    {
+      "epoch": 0.39836087909531237,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 7812
+    },
+    {
+      "epoch": 0.39841187255141775,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 7813
+    },
+    {
+      "epoch": 0.3984628660075231,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 7814
+    },
+    {
+      "epoch": 0.3985138594636285,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3985648529197339,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3986158463758393,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 7817
+    },
+    {
+      "epoch": 0.3986668398319447,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 7818
+    },
+    {
+      "epoch": 0.3987178332880501,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 7819
+    },
+    {
+      "epoch": 0.3987688267441555,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 7820
+    },
+    {
+      "epoch": 0.39881982020026086,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 7821
+    },
+    {
+      "epoch": 0.39887081365636623,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 7822
+    },
+    {
+      "epoch": 0.39892180711247166,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 7823
+    },
+    {
+      "epoch": 0.39897280056857704,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 7824
+    },
+    {
+      "epoch": 0.3990237940246824,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 7825
+    },
+    {
+      "epoch": 0.39907478748078784,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 7826
+    },
+    {
+      "epoch": 0.3991257809368932,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 7827
+    },
+    {
+      "epoch": 0.3991767743929986,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 7828
+    },
+    {
+      "epoch": 0.39922776784910396,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 7829
+    },
+    {
+      "epoch": 0.3992787613052094,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 7830
+    },
+    {
+      "epoch": 0.39932975476131477,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 7831
+    },
+    {
+      "epoch": 0.39938074821742015,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3994317416735256,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 7833
+    },
+    {
+      "epoch": 0.39948273512963095,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 7834
+    },
+    {
+      "epoch": 0.3995337285857363,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3995847220418417,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 7836
+    },
+    {
+      "epoch": 0.39963571549794713,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 7837
+    },
+    {
+      "epoch": 0.3996867089540525,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 7838
+    },
+    {
+      "epoch": 0.3997377024101579,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7839
+    },
+    {
+      "epoch": 0.3997886958662633,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 7840
+    },
+    {
+      "epoch": 0.3998396893223687,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 7841
+    },
+    {
+      "epoch": 0.39989068277847406,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 7842
+    },
+    {
+      "epoch": 0.39994167623457944,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 7843
+    },
+    {
+      "epoch": 0.39999266969068487,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 7844
+    },
+    {
+      "epoch": 0.40004366314679024,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 7845
+    },
+    {
+      "epoch": 0.4000946566028956,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 7846
+    },
+    {
+      "epoch": 0.40014565005900105,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 7847
+    },
+    {
+      "epoch": 0.4001966435151064,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 7848
+    },
+    {
+      "epoch": 0.4002476369712118,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 7849
+    },
+    {
+      "epoch": 0.40029863042731717,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4003496238834226,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 7851
+    },
+    {
+      "epoch": 0.400400617339528,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 7852
+    },
+    {
+      "epoch": 0.40045161079563335,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4005026042517388,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 7854
+    },
+    {
+      "epoch": 0.40055359770784416,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 7855
+    },
+    {
+      "epoch": 0.40060459116394953,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4006555846200549,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7857
+    },
+    {
+      "epoch": 0.40070657807616034,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 7858
+    },
+    {
+      "epoch": 0.4007575715322657,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 7859
+    },
+    {
+      "epoch": 0.4008085649883711,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4008595584444765,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 7861
+    },
+    {
+      "epoch": 0.4009105519005819,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 7862
+    },
+    {
+      "epoch": 0.40096154535668727,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 7863
+    },
+    {
+      "epoch": 0.40101253881279264,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 7864
+    },
+    {
+      "epoch": 0.40106353226889807,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 7865
+    },
+    {
+      "epoch": 0.40111452572500345,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 7866
+    },
+    {
+      "epoch": 0.4011655191811088,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 7867
+    },
+    {
+      "epoch": 0.40121651263721425,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 7868
+    },
+    {
+      "epoch": 0.4012675060933196,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 7869
+    },
+    {
+      "epoch": 0.401318499549425,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4013694930055304,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 7871
+    },
+    {
+      "epoch": 0.4014204864616358,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 7872
+    },
+    {
+      "epoch": 0.4014714799177412,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 7873
+    },
+    {
+      "epoch": 0.40152247337384656,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 7874
+    },
+    {
+      "epoch": 0.401573466829952,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 7875
+    },
+    {
+      "epoch": 0.40162446028605736,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 7876
+    },
+    {
+      "epoch": 0.40167545374216274,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 7877
+    },
+    {
+      "epoch": 0.4017264471982681,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 7878
+    },
+    {
+      "epoch": 0.40177744065437354,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 7879
+    },
+    {
+      "epoch": 0.4018284341104789,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4018794275665843,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 7881
+    },
+    {
+      "epoch": 0.4019304210226897,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 7882
+    },
+    {
+      "epoch": 0.4019814144787951,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 7883
+    },
+    {
+      "epoch": 0.40203240793490047,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 7884
+    },
+    {
+      "epoch": 0.40208340139100585,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4021343948471113,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 7886
+    },
+    {
+      "epoch": 0.40218538830321665,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 7887
+    },
+    {
+      "epoch": 0.402236381759322,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8614,
+      "step": 7888
+    },
+    {
+      "epoch": 0.40228737521542746,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 7889
+    },
+    {
+      "epoch": 0.40233836867153283,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4023893621276382,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 7891
+    },
+    {
+      "epoch": 0.4024403555837436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 7892
+    },
+    {
+      "epoch": 0.402491349039849,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 7893
+    },
+    {
+      "epoch": 0.4025423424959544,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 7894
+    },
+    {
+      "epoch": 0.40259333595205976,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 7895
+    },
+    {
+      "epoch": 0.4026443294081652,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 7896
+    },
+    {
+      "epoch": 0.40269532286427057,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 7897
+    },
+    {
+      "epoch": 0.40274631632037594,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 7898
+    },
+    {
+      "epoch": 0.4027973097764813,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 7899
+    },
+    {
+      "epoch": 0.40284830323258675,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4028992966886921,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 7901
+    },
+    {
+      "epoch": 0.4029502901447975,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 7902
+    },
+    {
+      "epoch": 0.4030012836009029,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 7903
+    },
+    {
+      "epoch": 0.4030522770570083,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 7904
+    },
+    {
+      "epoch": 0.4031032705131137,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 7905
+    },
+    {
+      "epoch": 0.40315426396921905,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8574,
+      "step": 7906
+    },
+    {
+      "epoch": 0.4032052574253245,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 7907
+    },
+    {
+      "epoch": 0.40325625088142986,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 7908
+    },
+    {
+      "epoch": 0.40330724433753523,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 7909
+    },
+    {
+      "epoch": 0.40335823779364066,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 7910
+    },
+    {
+      "epoch": 0.40340923124974604,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 7911
+    },
+    {
+      "epoch": 0.4034602247058514,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 7912
+    },
+    {
+      "epoch": 0.4035112181619568,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 7913
+    },
+    {
+      "epoch": 0.4035622116180622,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 7914
+    },
+    {
+      "epoch": 0.4036132050741676,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 7915
+    },
+    {
+      "epoch": 0.40366419853027297,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 7916
+    },
+    {
+      "epoch": 0.4037151919863784,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 7917
+    },
+    {
+      "epoch": 0.40376618544248377,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 7918
+    },
+    {
+      "epoch": 0.40381717889858915,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 7919
+    },
+    {
+      "epoch": 0.4038681723546945,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 7920
+    },
+    {
+      "epoch": 0.40391916581079995,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4039701592669053,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 7922
+    },
+    {
+      "epoch": 0.4040211527230107,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 7923
+    },
+    {
+      "epoch": 0.40407214617911613,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 7924
+    },
+    {
+      "epoch": 0.4041231396352215,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 7925
+    },
+    {
+      "epoch": 0.4041741330913269,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 7926
+    },
+    {
+      "epoch": 0.40422512654743226,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 7927
+    },
+    {
+      "epoch": 0.4042761200035377,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 7928
+    },
+    {
+      "epoch": 0.40432711345964306,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 7929
+    },
+    {
+      "epoch": 0.40437810691574844,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 7930
+    },
+    {
+      "epoch": 0.40442910037185387,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 7931
+    },
+    {
+      "epoch": 0.40448009382795924,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 7932
+    },
+    {
+      "epoch": 0.4045310872840646,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 7933
+    },
+    {
+      "epoch": 0.40458208074017,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 7934
+    },
+    {
+      "epoch": 0.4046330741962754,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 7935
+    },
+    {
+      "epoch": 0.4046840676523808,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 7936
+    },
+    {
+      "epoch": 0.40473506110848617,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8538,
+      "step": 7937
+    },
+    {
+      "epoch": 0.4047860545645916,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 7938
+    },
+    {
+      "epoch": 0.404837048020697,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 7939
+    },
+    {
+      "epoch": 0.40488804147680235,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4049390349329077,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7799,
+      "step": 7941
+    },
+    {
+      "epoch": 0.40499002838901316,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 7942
+    },
+    {
+      "epoch": 0.40504102184511853,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 7943
+    },
+    {
+      "epoch": 0.4050920153012239,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 7944
+    },
+    {
+      "epoch": 0.40514300875732934,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 7945
+    },
+    {
+      "epoch": 0.4051940022134347,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 7946
+    },
+    {
+      "epoch": 0.4052449956695401,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 7947
+    },
+    {
+      "epoch": 0.40529598912564546,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 7948
+    },
+    {
+      "epoch": 0.4053469825817509,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 7949
+    },
+    {
+      "epoch": 0.40539797603785627,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 7950
+    },
+    {
+      "epoch": 0.40544896949396164,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 7951
+    },
+    {
+      "epoch": 0.40549996295006707,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 7952
+    },
+    {
+      "epoch": 0.40555095640617245,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 7953
+    },
+    {
+      "epoch": 0.4056019498622778,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 7954
+    },
+    {
+      "epoch": 0.4056529433183832,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 7955
+    },
+    {
+      "epoch": 0.4057039367744886,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 7956
+    },
+    {
+      "epoch": 0.405754930230594,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 7957
+    },
+    {
+      "epoch": 0.4058059236866994,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 7958
+    },
+    {
+      "epoch": 0.4058569171428048,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 7959
+    },
+    {
+      "epoch": 0.4059079105989102,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 7960
+    },
+    {
+      "epoch": 0.40595890405501556,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 7961
+    },
+    {
+      "epoch": 0.40600989751112093,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 7962
+    },
+    {
+      "epoch": 0.40606089096722636,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 7963
+    },
+    {
+      "epoch": 0.40611188442333174,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 7964
+    },
+    {
+      "epoch": 0.4061628778794371,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 7965
+    },
+    {
+      "epoch": 0.40621387133554254,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 7966
+    },
+    {
+      "epoch": 0.4062648647916479,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 7967
+    },
+    {
+      "epoch": 0.4063158582477533,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 7968
+    },
+    {
+      "epoch": 0.40636685170385867,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 7969
+    },
+    {
+      "epoch": 0.4064178451599641,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 7970
+    },
+    {
+      "epoch": 0.40646883861606947,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 7971
+    },
+    {
+      "epoch": 0.40651983207217485,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 7972
+    },
+    {
+      "epoch": 0.4065708255282803,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 7973
+    },
+    {
+      "epoch": 0.40662181898438565,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 7974
+    },
+    {
+      "epoch": 0.406672812440491,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 7975
+    },
+    {
+      "epoch": 0.4067238058965964,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 7976
+    },
+    {
+      "epoch": 0.40677479935270183,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 7977
+    },
+    {
+      "epoch": 0.4068257928088072,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 7978
+    },
+    {
+      "epoch": 0.4068767862649126,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 7979
+    },
+    {
+      "epoch": 0.406927779721018,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4069787731771234,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8766,
+      "step": 7981
+    },
+    {
+      "epoch": 0.40702976663322876,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 7982
+    },
+    {
+      "epoch": 0.40708076008933414,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 7983
+    },
+    {
+      "epoch": 0.40713175354543957,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 7984
+    },
+    {
+      "epoch": 0.40718274700154494,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 7985
+    },
+    {
+      "epoch": 0.4072337404576503,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 7986
+    },
+    {
+      "epoch": 0.40728473391375575,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 7987
+    },
+    {
+      "epoch": 0.4073357273698611,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 7988
+    },
+    {
+      "epoch": 0.4073867208259665,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 7989
+    },
+    {
+      "epoch": 0.40743771428207187,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8462,
+      "step": 7990
+    },
+    {
+      "epoch": 0.4074887077381773,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 7991
+    },
+    {
+      "epoch": 0.4075397011942827,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 7992
+    },
+    {
+      "epoch": 0.40759069465038805,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 7993
+    },
+    {
+      "epoch": 0.4076416881064935,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 7994
+    },
+    {
+      "epoch": 0.40769268156259886,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 7995
+    },
+    {
+      "epoch": 0.40774367501870423,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 7996
+    },
+    {
+      "epoch": 0.4077946684748096,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 7997
+    },
+    {
+      "epoch": 0.40784566193091504,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 7998
+    },
+    {
+      "epoch": 0.4078966553870204,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 7999
+    },
+    {
+      "epoch": 0.4079476488431258,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8536,
+      "step": 8000
+    },
+    {
+      "epoch": 0.4079986422992312,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 8001
+    },
+    {
+      "epoch": 0.4080496357553366,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 8002
+    },
+    {
+      "epoch": 0.40810062921144197,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 8003
+    },
+    {
+      "epoch": 0.40815162266754734,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 8004
+    },
+    {
+      "epoch": 0.4082026161236528,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 8005
+    },
+    {
+      "epoch": 0.40825360957975815,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 8006
+    },
+    {
+      "epoch": 0.4083046030358635,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 8007
+    },
+    {
+      "epoch": 0.40835559649196895,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 8008
+    },
+    {
+      "epoch": 0.40840658994807433,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8523,
+      "step": 8009
+    },
+    {
+      "epoch": 0.4084575834041797,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4085085768602851,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 8011
+    },
+    {
+      "epoch": 0.4085595703163905,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 8012
+    },
+    {
+      "epoch": 0.4086105637724959,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 8013
+    },
+    {
+      "epoch": 0.40866155722860126,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 8014
+    },
+    {
+      "epoch": 0.4087125506847067,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 8015
+    },
+    {
+      "epoch": 0.40876354414081206,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 8016
+    },
+    {
+      "epoch": 0.40881453759691744,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 8017
+    },
+    {
+      "epoch": 0.4088655310530228,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 8018
+    },
+    {
+      "epoch": 0.40891652450912824,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 8019
+    },
+    {
+      "epoch": 0.4089675179652336,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 8020
+    },
+    {
+      "epoch": 0.409018511421339,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 8021
+    },
+    {
+      "epoch": 0.4090695048774444,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 8022
+    },
+    {
+      "epoch": 0.4091204983335498,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 8023
+    },
+    {
+      "epoch": 0.4091714917896552,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 8024
+    },
+    {
+      "epoch": 0.40922248524576055,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 8025
+    },
+    {
+      "epoch": 0.409273478701866,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 8026
+    },
+    {
+      "epoch": 0.40932447215797135,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 8027
+    },
+    {
+      "epoch": 0.40937546561407673,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 8028
+    },
+    {
+      "epoch": 0.40942645907018216,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 8029
+    },
+    {
+      "epoch": 0.40947745252628753,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4095284459823929,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 8031
+    },
+    {
+      "epoch": 0.4095794394384983,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 8032
+    },
+    {
+      "epoch": 0.4096304328946037,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 8033
+    },
+    {
+      "epoch": 0.4096814263507091,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 8034
+    },
+    {
+      "epoch": 0.40973241980681446,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4097834132629199,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 8036
+    },
+    {
+      "epoch": 0.40983440671902527,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 8037
+    },
+    {
+      "epoch": 0.40988540017513064,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 8038
+    },
+    {
+      "epoch": 0.409936393631236,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 8039
+    },
+    {
+      "epoch": 0.40998738708734145,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8040
+    },
+    {
+      "epoch": 0.4100383805434468,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 8041
+    },
+    {
+      "epoch": 0.4100893739995522,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 8042
+    },
+    {
+      "epoch": 0.41014036745565763,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 8043
+    },
+    {
+      "epoch": 0.410191360911763,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 8044
+    },
+    {
+      "epoch": 0.4102423543678684,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 8045
+    },
+    {
+      "epoch": 0.41029334782397375,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 8046
+    },
+    {
+      "epoch": 0.4103443412800792,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 8047
+    },
+    {
+      "epoch": 0.41039533473618456,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 8048
+    },
+    {
+      "epoch": 0.41044632819228993,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 8049
+    },
+    {
+      "epoch": 0.41049732164839536,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8471,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41054831510450074,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 8051
+    },
+    {
+      "epoch": 0.4105993085606061,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 8052
+    },
+    {
+      "epoch": 0.4106503020167115,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 8053
+    },
+    {
+      "epoch": 0.4107012954728169,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 8054
+    },
+    {
+      "epoch": 0.4107522889289223,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 8055
+    },
+    {
+      "epoch": 0.41080328238502767,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 8056
+    },
+    {
+      "epoch": 0.4108542758411331,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 8057
+    },
+    {
+      "epoch": 0.4109052692972385,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 8058
+    },
+    {
+      "epoch": 0.41095626275334385,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.851,
+      "step": 8059
+    },
+    {
+      "epoch": 0.4110072562094492,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41105824966555465,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 8061
+    },
+    {
+      "epoch": 0.41110924312166003,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4111602365777654,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 8063
+    },
+    {
+      "epoch": 0.41121123003387083,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4112622234899762,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 8065
+    },
+    {
+      "epoch": 0.4113132169460816,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 8066
+    },
+    {
+      "epoch": 0.41136421040218696,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 8067
+    },
+    {
+      "epoch": 0.4114152038582924,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 8068
+    },
+    {
+      "epoch": 0.41146619731439776,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8581,
+      "step": 8069
+    },
+    {
+      "epoch": 0.41151719077050314,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41156818422660857,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 8071
+    },
+    {
+      "epoch": 0.41161917768271394,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4116701711388193,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 8073
+    },
+    {
+      "epoch": 0.4117211645949247,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 8074
+    },
+    {
+      "epoch": 0.4117721580510301,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 8075
+    },
+    {
+      "epoch": 0.4118231515071355,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4118741449632409,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 8077
+    },
+    {
+      "epoch": 0.4119251384193463,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 8078
+    },
+    {
+      "epoch": 0.4119761318754517,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 8079
+    },
+    {
+      "epoch": 0.41202712533155705,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 8080
+    },
+    {
+      "epoch": 0.41207811878766243,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8081
+    },
+    {
+      "epoch": 0.41212911224376786,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 8082
+    },
+    {
+      "epoch": 0.41218010569987323,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 8083
+    },
+    {
+      "epoch": 0.4122310991559786,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 8084
+    },
+    {
+      "epoch": 0.41228209261208404,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 8085
+    },
+    {
+      "epoch": 0.4123330860681894,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8086
+    },
+    {
+      "epoch": 0.4123840795242948,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8087
+    },
+    {
+      "epoch": 0.41243507298040016,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8088
+    },
+    {
+      "epoch": 0.4124860664365056,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 8089
+    },
+    {
+      "epoch": 0.41253705989261097,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 8090
+    },
+    {
+      "epoch": 0.41258805334871634,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 8091
+    },
+    {
+      "epoch": 0.4126390468048218,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 8092
+    },
+    {
+      "epoch": 0.41269004026092715,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 8093
+    },
+    {
+      "epoch": 0.4127410337170325,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 8094
+    },
+    {
+      "epoch": 0.4127920271731379,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8095
+    },
+    {
+      "epoch": 0.41284302062924333,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 8096
+    },
+    {
+      "epoch": 0.4128940140853487,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 8097
+    },
+    {
+      "epoch": 0.4129450075414541,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 8098
+    },
+    {
+      "epoch": 0.4129960009975595,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 8099
+    },
+    {
+      "epoch": 0.4130469944536649,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8100
+    },
+    {
+      "epoch": 0.41309798790977026,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 8101
+    },
+    {
+      "epoch": 0.41314898136587563,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 8102
+    },
+    {
+      "epoch": 0.41319997482198106,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 8103
+    },
+    {
+      "epoch": 0.41325096827808644,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 8104
+    },
+    {
+      "epoch": 0.4133019617341918,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 8105
+    },
+    {
+      "epoch": 0.41335295519029724,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4134039486464026,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 8107
+    },
+    {
+      "epoch": 0.413454942102508,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8482,
+      "step": 8108
+    },
+    {
+      "epoch": 0.41350593555861337,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8109
+    },
+    {
+      "epoch": 0.4135569290147188,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8507,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4136079224708242,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 8111
+    },
+    {
+      "epoch": 0.41365891592692955,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 8112
+    },
+    {
+      "epoch": 0.413709909383035,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 8113
+    },
+    {
+      "epoch": 0.41376090283914035,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 8114
+    },
+    {
+      "epoch": 0.41381189629524573,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8115
+    },
+    {
+      "epoch": 0.4138628897513511,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 8116
+    },
+    {
+      "epoch": 0.41391388320745653,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 8117
+    },
+    {
+      "epoch": 0.4139648766635619,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 8118
+    },
+    {
+      "epoch": 0.4140158701196673,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 8119
+    },
+    {
+      "epoch": 0.4140668635757727,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4141178570318781,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 8121
+    },
+    {
+      "epoch": 0.41416885048798346,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 8122
+    },
+    {
+      "epoch": 0.41421984394408884,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 8123
+    },
+    {
+      "epoch": 0.41427083740019427,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 8124
+    },
+    {
+      "epoch": 0.41432183085629964,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 8125
+    },
+    {
+      "epoch": 0.414372824312405,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 8126
+    },
+    {
+      "epoch": 0.41442381776851045,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 8127
+    },
+    {
+      "epoch": 0.4144748112246158,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 8128
+    },
+    {
+      "epoch": 0.4145258046807212,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 8129
+    },
+    {
+      "epoch": 0.4145767981368266,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8547,
+      "step": 8130
+    },
+    {
+      "epoch": 0.414627791592932,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 8131
+    },
+    {
+      "epoch": 0.4146787850490374,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 8132
+    },
+    {
+      "epoch": 0.41472977850514275,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 8133
+    },
+    {
+      "epoch": 0.4147807719612482,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8577,
+      "step": 8134
+    },
+    {
+      "epoch": 0.41483176541735356,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8135
+    },
+    {
+      "epoch": 0.41488275887345893,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8136
+    },
+    {
+      "epoch": 0.4149337523295643,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8137
+    },
+    {
+      "epoch": 0.41498474578566974,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 8138
+    },
+    {
+      "epoch": 0.4150357392417751,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 8139
+    },
+    {
+      "epoch": 0.4150867326978805,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4151377261539859,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 8141
+    },
+    {
+      "epoch": 0.4151887196100913,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 8142
+    },
+    {
+      "epoch": 0.41523971306619667,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 8143
+    },
+    {
+      "epoch": 0.41529070652230204,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 8144
+    },
+    {
+      "epoch": 0.4153416999784075,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 8145
+    },
+    {
+      "epoch": 0.41539269343451285,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 8146
+    },
+    {
+      "epoch": 0.4154436868906182,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 8147
+    },
+    {
+      "epoch": 0.41549468034672365,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 8148
+    },
+    {
+      "epoch": 0.41554567380282903,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 8149
+    },
+    {
+      "epoch": 0.4155966672589344,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4156476607150398,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 8151
+    },
+    {
+      "epoch": 0.4156986541711452,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 8152
+    },
+    {
+      "epoch": 0.4157496476272506,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 8153
+    },
+    {
+      "epoch": 0.41580064108335596,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8559,
+      "step": 8154
+    },
+    {
+      "epoch": 0.4158516345394614,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 8155
+    },
+    {
+      "epoch": 0.41590262799556676,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.857,
+      "step": 8156
+    },
+    {
+      "epoch": 0.41595362145167214,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 8157
+    },
+    {
+      "epoch": 0.4160046149077775,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 8158
+    },
+    {
+      "epoch": 0.41605560836388294,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 8159
+    },
+    {
+      "epoch": 0.4161066018199883,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 8160
+    },
+    {
+      "epoch": 0.4161575952760937,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 8161
+    },
+    {
+      "epoch": 0.4162085887321991,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 8162
+    },
+    {
+      "epoch": 0.4162595821883045,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 8163
+    },
+    {
+      "epoch": 0.4163105756444099,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 8164
+    },
+    {
+      "epoch": 0.41636156910051525,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 8165
+    },
+    {
+      "epoch": 0.4164125625566207,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 8166
+    },
+    {
+      "epoch": 0.41646355601272605,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 8167
+    },
+    {
+      "epoch": 0.41651454946883143,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 8168
+    },
+    {
+      "epoch": 0.41656554292493686,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 8169
+    },
+    {
+      "epoch": 0.41661653638104224,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4166675298371476,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 8171
+    },
+    {
+      "epoch": 0.416718523293253,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4167695167493584,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 8173
+    },
+    {
+      "epoch": 0.4168205102054638,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 8174
+    },
+    {
+      "epoch": 0.41687150366156916,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4169224971176746,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 8176
+    },
+    {
+      "epoch": 0.41697349057377997,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 8177
+    },
+    {
+      "epoch": 0.41702448402988535,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 8178
+    },
+    {
+      "epoch": 0.4170754774859907,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 8179
+    },
+    {
+      "epoch": 0.41712647094209615,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4171774643982015,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 8181
+    },
+    {
+      "epoch": 0.4172284578543069,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 8182
+    },
+    {
+      "epoch": 0.41727945131041233,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 8183
+    },
+    {
+      "epoch": 0.4173304447665177,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 8184
+    },
+    {
+      "epoch": 0.4173814382226231,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 8185
+    },
+    {
+      "epoch": 0.41743243167872846,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 8186
+    },
+    {
+      "epoch": 0.4174834251348339,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 8187
+    },
+    {
+      "epoch": 0.41753441859093926,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 8188
+    },
+    {
+      "epoch": 0.41758541204704464,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 8189
+    },
+    {
+      "epoch": 0.41763640550315007,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 8190
+    },
+    {
+      "epoch": 0.41768739895925544,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 8191
+    },
+    {
+      "epoch": 0.4177383924153608,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 8192
+    },
+    {
+      "epoch": 0.4177893858714662,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 8193
+    },
+    {
+      "epoch": 0.4178403793275716,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 8194
+    },
+    {
+      "epoch": 0.417891372783677,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 8195
+    },
+    {
+      "epoch": 0.41794236623978237,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 8196
+    },
+    {
+      "epoch": 0.4179933596958878,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 8197
+    },
+    {
+      "epoch": 0.4180443531519932,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 8198
+    },
+    {
+      "epoch": 0.41809534660809855,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 8199
+    },
+    {
+      "epoch": 0.4181463400642039,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 8200
+    },
+    {
+      "epoch": 0.41819733352030936,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 8201
+    },
+    {
+      "epoch": 0.41824832697641473,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 8202
+    },
+    {
+      "epoch": 0.4182993204325201,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 8203
+    },
+    {
+      "epoch": 0.41835031388862554,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 8204
+    },
+    {
+      "epoch": 0.4184013073447309,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8205
+    },
+    {
+      "epoch": 0.4184523008008363,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 8206
+    },
+    {
+      "epoch": 0.41850329425694166,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 8207
+    },
+    {
+      "epoch": 0.4185542877130471,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 8208
+    },
+    {
+      "epoch": 0.41860528116915247,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 8209
+    },
+    {
+      "epoch": 0.41865627462525784,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 8210
+    },
+    {
+      "epoch": 0.41870726808136327,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 8211
+    },
+    {
+      "epoch": 0.41875826153746865,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 8212
+    },
+    {
+      "epoch": 0.418809254993574,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 8213
+    },
+    {
+      "epoch": 0.4188602484496794,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 8214
+    },
+    {
+      "epoch": 0.4189112419057848,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 8215
+    },
+    {
+      "epoch": 0.4189622353618902,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 8216
+    },
+    {
+      "epoch": 0.4190132288179956,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 8217
+    },
+    {
+      "epoch": 0.419064222274101,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 8218
+    },
+    {
+      "epoch": 0.4191152157302064,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 8219
+    },
+    {
+      "epoch": 0.41916620918631176,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 8220
+    },
+    {
+      "epoch": 0.41921720264241713,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 8221
+    },
+    {
+      "epoch": 0.41926819609852256,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 8222
+    },
+    {
+      "epoch": 0.41931918955462794,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 8223
+    },
+    {
+      "epoch": 0.4193701830107333,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 8224
+    },
+    {
+      "epoch": 0.41942117646683874,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 8225
+    },
+    {
+      "epoch": 0.4194721699229441,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 8226
+    },
+    {
+      "epoch": 0.4195231633790495,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 8227
+    },
+    {
+      "epoch": 0.41957415683515487,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 8228
+    },
+    {
+      "epoch": 0.4196251502912603,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 8229
+    },
+    {
+      "epoch": 0.41967614374736567,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 8230
+    },
+    {
+      "epoch": 0.41972713720347105,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 8231
+    },
+    {
+      "epoch": 0.4197781306595765,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 8232
+    },
+    {
+      "epoch": 0.41982912411568185,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 8233
+    },
+    {
+      "epoch": 0.4198801175717872,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 8234
+    },
+    {
+      "epoch": 0.4199311110278926,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 8235
+    },
+    {
+      "epoch": 0.41998210448399803,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 8236
+    },
+    {
+      "epoch": 0.4200330979401034,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 8237
+    },
+    {
+      "epoch": 0.4200840913962088,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 8238
+    },
+    {
+      "epoch": 0.4201350848523142,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 8239
+    },
+    {
+      "epoch": 0.4201860783084196,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 8240
+    },
+    {
+      "epoch": 0.42023707176452496,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 8241
+    },
+    {
+      "epoch": 0.42028806522063034,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 8242
+    },
+    {
+      "epoch": 0.42033905867673577,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 8243
+    },
+    {
+      "epoch": 0.42039005213284114,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 8244
+    },
+    {
+      "epoch": 0.4204410455889465,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 8245
+    },
+    {
+      "epoch": 0.42049203904505195,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 8246
+    },
+    {
+      "epoch": 0.4205430325011573,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8247
+    },
+    {
+      "epoch": 0.4205940259572627,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 8248
+    },
+    {
+      "epoch": 0.42064501941336807,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 8249
+    },
+    {
+      "epoch": 0.4206960128694735,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4207470063255789,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 8251
+    },
+    {
+      "epoch": 0.42079799978168425,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8252
+    },
+    {
+      "epoch": 0.4208489932377897,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 8253
+    },
+    {
+      "epoch": 0.42089998669389506,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 8254
+    },
+    {
+      "epoch": 0.42095098015000043,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 8255
+    },
+    {
+      "epoch": 0.4210019736061058,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 8256
+    },
+    {
+      "epoch": 0.42105296706221124,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 8257
+    },
+    {
+      "epoch": 0.4211039605183166,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 8258
+    },
+    {
+      "epoch": 0.421154953974422,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 8259
+    },
+    {
+      "epoch": 0.4212059474305274,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4212569408866328,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8261
+    },
+    {
+      "epoch": 0.42130793434273817,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 8262
+    },
+    {
+      "epoch": 0.42135892779884354,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 8263
+    },
+    {
+      "epoch": 0.42140992125494897,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 8264
+    },
+    {
+      "epoch": 0.42146091471105435,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 8265
+    },
+    {
+      "epoch": 0.4215119081671597,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 8266
+    },
+    {
+      "epoch": 0.42156290162326515,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7764,
+      "step": 8267
+    },
+    {
+      "epoch": 0.4216138950793705,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 8268
+    },
+    {
+      "epoch": 0.4216648885354759,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 8269
+    },
+    {
+      "epoch": 0.4217158819915813,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 8270
+    },
+    {
+      "epoch": 0.4217668754476867,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 8271
+    },
+    {
+      "epoch": 0.4218178689037921,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.847,
+      "step": 8272
+    },
+    {
+      "epoch": 0.42186886235989746,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 8273
+    },
+    {
+      "epoch": 0.4219198558160029,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 8274
+    },
+    {
+      "epoch": 0.42197084927210826,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 8275
+    },
+    {
+      "epoch": 0.42202184272821364,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 8276
+    },
+    {
+      "epoch": 0.422072836184319,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 8277
+    },
+    {
+      "epoch": 0.42212382964042444,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 8278
+    },
+    {
+      "epoch": 0.4221748230965298,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 8279
+    },
+    {
+      "epoch": 0.4222258165526352,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4222768100087406,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 8281
+    },
+    {
+      "epoch": 0.422327803464846,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 8282
+    },
+    {
+      "epoch": 0.42237879692095137,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 8283
+    },
+    {
+      "epoch": 0.42242979037705675,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 8284
+    },
+    {
+      "epoch": 0.4224807838331622,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 8285
+    },
+    {
+      "epoch": 0.42253177728926755,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 8286
+    },
+    {
+      "epoch": 0.4225827707453729,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 8287
+    },
+    {
+      "epoch": 0.42263376420147836,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 8288
+    },
+    {
+      "epoch": 0.42268475765758373,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 8289
+    },
+    {
+      "epoch": 0.4227357511136891,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4227867445697945,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 8291
+    },
+    {
+      "epoch": 0.4228377380258999,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 8292
+    },
+    {
+      "epoch": 0.4228887314820053,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 8293
+    },
+    {
+      "epoch": 0.42293972493811066,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 8294
+    },
+    {
+      "epoch": 0.4229907183942161,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 8295
+    },
+    {
+      "epoch": 0.42304171185032147,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 8296
+    },
+    {
+      "epoch": 0.42309270530642684,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 8297
+    },
+    {
+      "epoch": 0.4231436987625322,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 8298
+    },
+    {
+      "epoch": 0.42319469221863765,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8299
+    },
+    {
+      "epoch": 0.423245685674743,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4232966791308484,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 8301
+    },
+    {
+      "epoch": 0.4233476725869538,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 8302
+    },
+    {
+      "epoch": 0.4233986660430592,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 8303
+    },
+    {
+      "epoch": 0.4234496594991646,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 8304
+    },
+    {
+      "epoch": 0.42350065295526995,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 8305
+    },
+    {
+      "epoch": 0.4235516464113754,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 8306
+    },
+    {
+      "epoch": 0.42360263986748076,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8508,
+      "step": 8307
+    },
+    {
+      "epoch": 0.42365363332358613,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 8308
+    },
+    {
+      "epoch": 0.42370462677969156,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8309
+    },
+    {
+      "epoch": 0.42375562023579694,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4238066136919023,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 8311
+    },
+    {
+      "epoch": 0.4238576071480077,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 8312
+    },
+    {
+      "epoch": 0.4239086006041131,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 8313
+    },
+    {
+      "epoch": 0.4239595940602185,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 8314
+    },
+    {
+      "epoch": 0.42401058751632387,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 8315
+    },
+    {
+      "epoch": 0.4240615809724293,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 8316
+    },
+    {
+      "epoch": 0.42411257442853467,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 8317
+    },
+    {
+      "epoch": 0.42416356788464005,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 8318
+    },
+    {
+      "epoch": 0.4242145613407454,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 8319
+    },
+    {
+      "epoch": 0.42426555479685085,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4243165482529562,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 8321
+    },
+    {
+      "epoch": 0.4243675417090616,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 8322
+    },
+    {
+      "epoch": 0.42441853516516703,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 8323
+    },
+    {
+      "epoch": 0.4244695286212724,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 8324
+    },
+    {
+      "epoch": 0.4245205220773778,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 8325
+    },
+    {
+      "epoch": 0.42457151553348316,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8326
+    },
+    {
+      "epoch": 0.4246225089895886,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 8327
+    },
+    {
+      "epoch": 0.42467350244569396,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 8328
+    },
+    {
+      "epoch": 0.42472449590179934,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 8329
+    },
+    {
+      "epoch": 0.42477548935790477,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42482648281401014,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 8331
+    },
+    {
+      "epoch": 0.4248774762701155,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 8332
+    },
+    {
+      "epoch": 0.4249284697262209,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 8333
+    },
+    {
+      "epoch": 0.4249794631823263,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 8334
+    },
+    {
+      "epoch": 0.4250304566384317,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 8335
+    },
+    {
+      "epoch": 0.42508145009453707,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 8336
+    },
+    {
+      "epoch": 0.4251324435506425,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 8337
+    },
+    {
+      "epoch": 0.4251834370067479,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 8338
+    },
+    {
+      "epoch": 0.42523443046285325,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 8339
+    },
+    {
+      "epoch": 0.4252854239189586,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 8340
+    },
+    {
+      "epoch": 0.42533641737506406,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 8341
+    },
+    {
+      "epoch": 0.42538741083116943,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 8342
+    },
+    {
+      "epoch": 0.4254384042872748,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 8343
+    },
+    {
+      "epoch": 0.42548939774338024,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 8344
+    },
+    {
+      "epoch": 0.4255403911994856,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 8345
+    },
+    {
+      "epoch": 0.425591384655591,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 8346
+    },
+    {
+      "epoch": 0.42564237811169636,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 8347
+    },
+    {
+      "epoch": 0.4256933715678018,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 8348
+    },
+    {
+      "epoch": 0.42574436502390717,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 8349
+    },
+    {
+      "epoch": 0.42579535848001254,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 8350
+    },
+    {
+      "epoch": 0.425846351936118,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 8351
+    },
+    {
+      "epoch": 0.42589734539222335,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 8352
+    },
+    {
+      "epoch": 0.4259483388483287,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 8353
+    },
+    {
+      "epoch": 0.4259993323044341,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 8354
+    },
+    {
+      "epoch": 0.4260503257605395,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 8355
+    },
+    {
+      "epoch": 0.4261013192166449,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 8356
+    },
+    {
+      "epoch": 0.4261523126727503,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8357
+    },
+    {
+      "epoch": 0.4262033061288557,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8466,
+      "step": 8358
+    },
+    {
+      "epoch": 0.4262542995849611,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 8359
+    },
+    {
+      "epoch": 0.42630529304106646,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 8360
+    },
+    {
+      "epoch": 0.42635628649717183,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 8361
+    },
+    {
+      "epoch": 0.42640727995327726,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 8362
+    },
+    {
+      "epoch": 0.42645827340938264,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 8363
+    },
+    {
+      "epoch": 0.426509266865488,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8364
+    },
+    {
+      "epoch": 0.42656026032159344,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8365
+    },
+    {
+      "epoch": 0.4266112537776988,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 8366
+    },
+    {
+      "epoch": 0.4266622472338042,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 8367
+    },
+    {
+      "epoch": 0.42671324068990957,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 8368
+    },
+    {
+      "epoch": 0.426764234146015,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 8369
+    },
+    {
+      "epoch": 0.4268152276021204,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 8370
+    },
+    {
+      "epoch": 0.42686622105822575,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8371
+    },
+    {
+      "epoch": 0.4269172145143312,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 8372
+    },
+    {
+      "epoch": 0.42696820797043655,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 8373
+    },
+    {
+      "epoch": 0.42701920142654193,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 8374
+    },
+    {
+      "epoch": 0.4270701948826473,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 8375
+    },
+    {
+      "epoch": 0.42712118833875273,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 8376
+    },
+    {
+      "epoch": 0.4271721817948581,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 8377
+    },
+    {
+      "epoch": 0.4272231752509635,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 8378
+    },
+    {
+      "epoch": 0.4272741687070689,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 8379
+    },
+    {
+      "epoch": 0.4273251621631743,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 8380
+    },
+    {
+      "epoch": 0.42737615561927966,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 8381
+    },
+    {
+      "epoch": 0.42742714907538504,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 8382
+    },
+    {
+      "epoch": 0.42747814253149047,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 8383
+    },
+    {
+      "epoch": 0.42752913598759584,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8384
+    },
+    {
+      "epoch": 0.4275801294437012,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 8385
+    },
+    {
+      "epoch": 0.42763112289980665,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 8386
+    },
+    {
+      "epoch": 0.427682116355912,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 8387
+    },
+    {
+      "epoch": 0.4277331098120174,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 8388
+    },
+    {
+      "epoch": 0.4277841032681228,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 8389
+    },
+    {
+      "epoch": 0.4278350967242282,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4278860901803336,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 8391
+    },
+    {
+      "epoch": 0.42793708363643895,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 8392
+    },
+    {
+      "epoch": 0.4279880770925444,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 8393
+    },
+    {
+      "epoch": 0.42803907054864976,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8472,
+      "step": 8394
+    },
+    {
+      "epoch": 0.42809006400475513,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8527,
+      "step": 8395
+    },
+    {
+      "epoch": 0.4281410574608605,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 8396
+    },
+    {
+      "epoch": 0.42819205091696594,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 8397
+    },
+    {
+      "epoch": 0.4282430443730713,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 8398
+    },
+    {
+      "epoch": 0.4282940378291767,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 8399
+    },
+    {
+      "epoch": 0.4283450312852821,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4283960247413875,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 8401
+    },
+    {
+      "epoch": 0.42844701819749287,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 8402
+    },
+    {
+      "epoch": 0.42849801165359824,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 8403
+    },
+    {
+      "epoch": 0.4285490051097037,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 8404
+    },
+    {
+      "epoch": 0.42859999856580905,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 8405
+    },
+    {
+      "epoch": 0.4286509920219144,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 8406
+    },
+    {
+      "epoch": 0.42870198547801985,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 8407
+    },
+    {
+      "epoch": 0.42875297893412523,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 8408
+    },
+    {
+      "epoch": 0.4288039723902306,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 8409
+    },
+    {
+      "epoch": 0.428854965846336,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4289059593024414,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 8411
+    },
+    {
+      "epoch": 0.4289569527585468,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 8412
+    },
+    {
+      "epoch": 0.42900794621465216,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 8413
+    },
+    {
+      "epoch": 0.4290589396707576,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 8414
+    },
+    {
+      "epoch": 0.42910993312686296,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 8415
+    },
+    {
+      "epoch": 0.42916092658296834,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 8416
+    },
+    {
+      "epoch": 0.4292119200390737,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 8417
+    },
+    {
+      "epoch": 0.42926291349517914,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 8418
+    },
+    {
+      "epoch": 0.4293139069512845,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8419
+    },
+    {
+      "epoch": 0.4293649004073899,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 8420
+    },
+    {
+      "epoch": 0.4294158938634953,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 8421
+    },
+    {
+      "epoch": 0.4294668873196007,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 8422
+    },
+    {
+      "epoch": 0.4295178807757061,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 8423
+    },
+    {
+      "epoch": 0.42956887423181145,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 8424
+    },
+    {
+      "epoch": 0.4296198676879169,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 8425
+    },
+    {
+      "epoch": 0.42967086114402225,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 8426
+    },
+    {
+      "epoch": 0.42972185460012763,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 8427
+    },
+    {
+      "epoch": 0.42977284805623306,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 8428
+    },
+    {
+      "epoch": 0.42982384151233843,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 8429
+    },
+    {
+      "epoch": 0.4298748349684438,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4299258284245492,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 8431
+    },
+    {
+      "epoch": 0.4299768218806546,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 8432
+    },
+    {
+      "epoch": 0.43002781533676,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 8433
+    },
+    {
+      "epoch": 0.43007880879286536,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7872,
+      "step": 8434
+    },
+    {
+      "epoch": 0.4301298022489708,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8435
+    },
+    {
+      "epoch": 0.43018079570507617,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.781,
+      "step": 8436
+    },
+    {
+      "epoch": 0.43023178916118154,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 8437
+    },
+    {
+      "epoch": 0.4302827826172869,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 8438
+    },
+    {
+      "epoch": 0.43033377607339235,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 8439
+    },
+    {
+      "epoch": 0.4303847695294977,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4304357629856031,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 8441
+    },
+    {
+      "epoch": 0.43048675644170853,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 8442
+    },
+    {
+      "epoch": 0.4305377498978139,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 8443
+    },
+    {
+      "epoch": 0.4305887433539193,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 8444
+    },
+    {
+      "epoch": 0.43063973681002465,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 8445
+    },
+    {
+      "epoch": 0.4306907302661301,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 8446
+    },
+    {
+      "epoch": 0.43074172372223546,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8447
+    },
+    {
+      "epoch": 0.43079271717834083,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 8448
+    },
+    {
+      "epoch": 0.43084371063444626,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 8449
+    },
+    {
+      "epoch": 0.43089470409055164,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 8450
+    },
+    {
+      "epoch": 0.430945697546657,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 8451
+    },
+    {
+      "epoch": 0.4309966910027624,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 8452
+    },
+    {
+      "epoch": 0.4310476844588678,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 8453
+    },
+    {
+      "epoch": 0.4310986779149732,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 8454
+    },
+    {
+      "epoch": 0.43114967137107857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 8455
+    },
+    {
+      "epoch": 0.431200664827184,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 8456
+    },
+    {
+      "epoch": 0.4312516582832894,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 8457
+    },
+    {
+      "epoch": 0.43130265173939475,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 8458
+    },
+    {
+      "epoch": 0.4313536451955001,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 8459
+    },
+    {
+      "epoch": 0.43140463865160555,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43145563210771093,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 8461
+    },
+    {
+      "epoch": 0.4315066255638163,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 8462
+    },
+    {
+      "epoch": 0.43155761901992173,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 8463
+    },
+    {
+      "epoch": 0.4316086124760271,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 8464
+    },
+    {
+      "epoch": 0.4316596059321325,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 8465
+    },
+    {
+      "epoch": 0.43171059938823786,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 8466
+    },
+    {
+      "epoch": 0.4317615928443433,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 8467
+    },
+    {
+      "epoch": 0.43181258630044866,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 8468
+    },
+    {
+      "epoch": 0.43186357975655404,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 8469
+    },
+    {
+      "epoch": 0.43191457321265947,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 8470
+    },
+    {
+      "epoch": 0.43196556666876484,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 8471
+    },
+    {
+      "epoch": 0.4320165601248702,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 8472
+    },
+    {
+      "epoch": 0.4320675535809756,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 8473
+    },
+    {
+      "epoch": 0.432118547037081,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 8474
+    },
+    {
+      "epoch": 0.4321695404931864,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 8475
+    },
+    {
+      "epoch": 0.4322205339492918,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 8476
+    },
+    {
+      "epoch": 0.4322715274053972,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 8477
+    },
+    {
+      "epoch": 0.4323225208615026,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 8478
+    },
+    {
+      "epoch": 0.43237351431760795,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 8479
+    },
+    {
+      "epoch": 0.43242450777371333,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 8480
+    },
+    {
+      "epoch": 0.43247550122981876,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 8481
+    },
+    {
+      "epoch": 0.43252649468592413,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8482
+    },
+    {
+      "epoch": 0.4325774881420295,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 8483
+    },
+    {
+      "epoch": 0.43262848159813494,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 8484
+    },
+    {
+      "epoch": 0.4326794750542403,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 8485
+    },
+    {
+      "epoch": 0.4327304685103457,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 8486
+    },
+    {
+      "epoch": 0.43278146196645106,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 8487
+    },
+    {
+      "epoch": 0.4328324554225565,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8488
+    },
+    {
+      "epoch": 0.43288344887866187,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 8489
+    },
+    {
+      "epoch": 0.43293444233476724,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4329854357908727,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8491
+    },
+    {
+      "epoch": 0.43303642924697805,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 8492
+    },
+    {
+      "epoch": 0.4330874227030834,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8493
+    },
+    {
+      "epoch": 0.4331384161591888,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 8494
+    },
+    {
+      "epoch": 0.43318940961529423,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 8495
+    },
+    {
+      "epoch": 0.4332404030713996,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 8496
+    },
+    {
+      "epoch": 0.433291396527505,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 8497
+    },
+    {
+      "epoch": 0.4333423899836104,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8498
+    },
+    {
+      "epoch": 0.4333933834397158,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 8499
+    },
+    {
+      "epoch": 0.43344437689582116,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 8500
+    },
+    {
+      "epoch": 0.43349537035192653,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 8501
+    },
+    {
+      "epoch": 0.43354636380803196,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 8502
+    },
+    {
+      "epoch": 0.43359735726413734,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 8503
+    },
+    {
+      "epoch": 0.4336483507202427,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 8504
+    },
+    {
+      "epoch": 0.43369934417634814,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 8505
+    },
+    {
+      "epoch": 0.4337503376324535,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8506
+    },
+    {
+      "epoch": 0.4338013310885589,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 8507
+    },
+    {
+      "epoch": 0.43385232454466427,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 8508
+    },
+    {
+      "epoch": 0.4339033180007697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 8509
+    },
+    {
+      "epoch": 0.4339543114568751,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43400530491298045,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 8511
+    },
+    {
+      "epoch": 0.4340562983690859,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 8512
+    },
+    {
+      "epoch": 0.43410729182519125,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 8513
+    },
+    {
+      "epoch": 0.43415828528129663,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 8514
+    },
+    {
+      "epoch": 0.434209278737402,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 8515
+    },
+    {
+      "epoch": 0.43426027219350744,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 8516
+    },
+    {
+      "epoch": 0.4343112656496128,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8435,
+      "step": 8517
+    },
+    {
+      "epoch": 0.4343622591057182,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 8518
+    },
+    {
+      "epoch": 0.4344132525618236,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 8519
+    },
+    {
+      "epoch": 0.434464246017929,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43451523947403436,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 8521
+    },
+    {
+      "epoch": 0.43456623293013974,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 8522
+    },
+    {
+      "epoch": 0.43461722638624517,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 8523
+    },
+    {
+      "epoch": 0.43466821984235054,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 8524
+    },
+    {
+      "epoch": 0.4347192132984559,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 8525
+    },
+    {
+      "epoch": 0.43477020675456135,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 8526
+    },
+    {
+      "epoch": 0.4348212002106667,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 8527
+    },
+    {
+      "epoch": 0.4348721936667721,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 8528
+    },
+    {
+      "epoch": 0.4349231871228775,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 8529
+    },
+    {
+      "epoch": 0.4349741805789829,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4350251740350883,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 8531
+    },
+    {
+      "epoch": 0.43507616749119365,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8532
+    },
+    {
+      "epoch": 0.4351271609472991,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 8533
+    },
+    {
+      "epoch": 0.43517815440340446,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 8534
+    },
+    {
+      "epoch": 0.43522914785950984,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4352801413156152,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 8536
+    },
+    {
+      "epoch": 0.43533113477172064,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 8537
+    },
+    {
+      "epoch": 0.435382128227826,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 8538
+    },
+    {
+      "epoch": 0.4354331216839314,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 8539
+    },
+    {
+      "epoch": 0.4354841151400368,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4355351085961422,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 8541
+    },
+    {
+      "epoch": 0.43558610205224757,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 8542
+    },
+    {
+      "epoch": 0.43563709550835295,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 8543
+    },
+    {
+      "epoch": 0.4356880889644584,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 8544
+    },
+    {
+      "epoch": 0.43573908242056375,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 8545
+    },
+    {
+      "epoch": 0.4357900758766691,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 8546
+    },
+    {
+      "epoch": 0.43584106933277456,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 8547
+    },
+    {
+      "epoch": 0.43589206278887993,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 8548
+    },
+    {
+      "epoch": 0.4359430562449853,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 8549
+    },
+    {
+      "epoch": 0.4359940497010907,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4360450431571961,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8551
+    },
+    {
+      "epoch": 0.4360960366133015,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 8552
+    },
+    {
+      "epoch": 0.43614703006940686,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 8553
+    },
+    {
+      "epoch": 0.4361980235255123,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 8554
+    },
+    {
+      "epoch": 0.43624901698161767,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 8555
+    },
+    {
+      "epoch": 0.43630001043772304,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 8556
+    },
+    {
+      "epoch": 0.4363510038938284,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 8557
+    },
+    {
+      "epoch": 0.43640199734993385,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 8558
+    },
+    {
+      "epoch": 0.4364529908060392,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 8559
+    },
+    {
+      "epoch": 0.4365039842621446,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 8560
+    },
+    {
+      "epoch": 0.43655497771825,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 8561
+    },
+    {
+      "epoch": 0.4366059711743554,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8562
+    },
+    {
+      "epoch": 0.4366569646304608,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 8563
+    },
+    {
+      "epoch": 0.43670795808656615,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 8564
+    },
+    {
+      "epoch": 0.4367589515426716,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 8565
+    },
+    {
+      "epoch": 0.43680994499877696,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 8566
+    },
+    {
+      "epoch": 0.43686093845488233,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 8567
+    },
+    {
+      "epoch": 0.43691193191098776,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 8568
+    },
+    {
+      "epoch": 0.43696292536709314,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8474,
+      "step": 8569
+    },
+    {
+      "epoch": 0.4370139188231985,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4370649122793039,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 8571
+    },
+    {
+      "epoch": 0.4371159057354093,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 8572
+    },
+    {
+      "epoch": 0.4371668991915147,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 8573
+    },
+    {
+      "epoch": 0.43721789264762007,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8574
+    },
+    {
+      "epoch": 0.4372688861037255,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 8575
+    },
+    {
+      "epoch": 0.43731987955983087,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 8576
+    },
+    {
+      "epoch": 0.43737087301593625,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 8577
+    },
+    {
+      "epoch": 0.4374218664720416,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 8578
+    },
+    {
+      "epoch": 0.43747285992814705,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 8579
+    },
+    {
+      "epoch": 0.4375238533842524,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4375748468403578,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8581
+    },
+    {
+      "epoch": 0.43762584029646323,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 8582
+    },
+    {
+      "epoch": 0.4376768337525686,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 8583
+    },
+    {
+      "epoch": 0.437727827208674,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 8584
+    },
+    {
+      "epoch": 0.43777882066477936,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 8585
+    },
+    {
+      "epoch": 0.4378298141208848,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 8586
+    },
+    {
+      "epoch": 0.43788080757699016,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7893,
+      "step": 8587
+    },
+    {
+      "epoch": 0.43793180103309554,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 8588
+    },
+    {
+      "epoch": 0.43798279448920097,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 8589
+    },
+    {
+      "epoch": 0.43803378794530634,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4380847814014117,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 8591
+    },
+    {
+      "epoch": 0.4381357748575171,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 8592
+    },
+    {
+      "epoch": 0.4381867683136225,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 8593
+    },
+    {
+      "epoch": 0.4382377617697279,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 8594
+    },
+    {
+      "epoch": 0.43828875522583327,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 8595
+    },
+    {
+      "epoch": 0.4383397486819387,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 8596
+    },
+    {
+      "epoch": 0.4383907421380441,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 8597
+    },
+    {
+      "epoch": 0.43844173559414945,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 8598
+    },
+    {
+      "epoch": 0.4384927290502548,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 8599
+    },
+    {
+      "epoch": 0.43854372250636026,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 8600
+    },
+    {
+      "epoch": 0.43859471596246563,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8601
+    },
+    {
+      "epoch": 0.438645709418571,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 8602
+    },
+    {
+      "epoch": 0.43869670287467644,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 8603
+    },
+    {
+      "epoch": 0.4387476963307818,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 8604
+    },
+    {
+      "epoch": 0.4387986897868872,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 8605
+    },
+    {
+      "epoch": 0.43884968324299256,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 8606
+    },
+    {
+      "epoch": 0.438900676699098,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 8607
+    },
+    {
+      "epoch": 0.43895167015520337,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 8608
+    },
+    {
+      "epoch": 0.43900266361130874,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 8609
+    },
+    {
+      "epoch": 0.43905365706741417,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 8610
+    },
+    {
+      "epoch": 0.43910465052351955,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 8611
+    },
+    {
+      "epoch": 0.4391556439796249,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 8612
+    },
+    {
+      "epoch": 0.4392066374357303,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 8613
+    },
+    {
+      "epoch": 0.4392576308918357,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 8614
+    },
+    {
+      "epoch": 0.4393086243479411,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8718,
+      "step": 8615
+    },
+    {
+      "epoch": 0.4393596178040465,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 8616
+    },
+    {
+      "epoch": 0.4394106112601519,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 8617
+    },
+    {
+      "epoch": 0.4394616047162573,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 8618
+    },
+    {
+      "epoch": 0.43951259817236266,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 8619
+    },
+    {
+      "epoch": 0.43956359162846803,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 8620
+    },
+    {
+      "epoch": 0.43961458508457346,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 8621
+    },
+    {
+      "epoch": 0.43966557854067884,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 8622
+    },
+    {
+      "epoch": 0.4397165719967842,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 8623
+    },
+    {
+      "epoch": 0.43976756545288964,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 8624
+    },
+    {
+      "epoch": 0.439818558908995,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 8625
+    },
+    {
+      "epoch": 0.4398695523651004,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 8626
+    },
+    {
+      "epoch": 0.43992054582120577,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 8627
+    },
+    {
+      "epoch": 0.4399715392773112,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 8628
+    },
+    {
+      "epoch": 0.44002253273341657,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 8629
+    },
+    {
+      "epoch": 0.44007352618952195,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4401245196456274,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8631
+    },
+    {
+      "epoch": 0.44017551310173275,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 8632
+    },
+    {
+      "epoch": 0.4402265065578381,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 8633
+    },
+    {
+      "epoch": 0.4402775000139435,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 8634
+    },
+    {
+      "epoch": 0.44032849347004893,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8635
+    },
+    {
+      "epoch": 0.4403794869261543,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 8636
+    },
+    {
+      "epoch": 0.4404304803822597,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 8637
+    },
+    {
+      "epoch": 0.4404814738383651,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 8638
+    },
+    {
+      "epoch": 0.4405324672944705,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 8639
+    },
+    {
+      "epoch": 0.44058346075057586,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 8640
+    },
+    {
+      "epoch": 0.44063445420668124,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 8641
+    },
+    {
+      "epoch": 0.44068544766278667,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 8642
+    },
+    {
+      "epoch": 0.44073644111889204,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 8643
+    },
+    {
+      "epoch": 0.4407874345749974,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 8644
+    },
+    {
+      "epoch": 0.44083842803110285,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 8645
+    },
+    {
+      "epoch": 0.4408894214872082,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 8646
+    },
+    {
+      "epoch": 0.4409404149433136,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 8647
+    },
+    {
+      "epoch": 0.44099140839941897,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 8648
+    },
+    {
+      "epoch": 0.4410424018555244,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 8649
+    },
+    {
+      "epoch": 0.4410933953116298,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 8650
+    },
+    {
+      "epoch": 0.44114438876773515,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4411953822238406,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 8652
+    },
+    {
+      "epoch": 0.44124637567994596,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 8653
+    },
+    {
+      "epoch": 0.44129736913605133,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 8654
+    },
+    {
+      "epoch": 0.4413483625921567,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 8655
+    },
+    {
+      "epoch": 0.44139935604826214,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 8656
+    },
+    {
+      "epoch": 0.4414503495043675,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4415013429604729,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 8658
+    },
+    {
+      "epoch": 0.4415523364165783,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 8659
+    },
+    {
+      "epoch": 0.4416033298726837,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 8660
+    },
+    {
+      "epoch": 0.44165432332878907,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 8661
+    },
+    {
+      "epoch": 0.44170531678489444,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 8662
+    },
+    {
+      "epoch": 0.44175631024099987,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 8663
+    },
+    {
+      "epoch": 0.44180730369710525,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 8664
+    },
+    {
+      "epoch": 0.4418582971532106,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 8665
+    },
+    {
+      "epoch": 0.44190929060931605,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 8666
+    },
+    {
+      "epoch": 0.4419602840654214,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4420112775215268,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 8668
+    },
+    {
+      "epoch": 0.4420622709776322,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 8669
+    },
+    {
+      "epoch": 0.4421132644337376,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7813,
+      "step": 8670
+    },
+    {
+      "epoch": 0.442164257889843,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 8671
+    },
+    {
+      "epoch": 0.44221525134594836,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 8672
+    },
+    {
+      "epoch": 0.4422662448020538,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 8673
+    },
+    {
+      "epoch": 0.44231723825815916,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 8674
+    },
+    {
+      "epoch": 0.44236823171426454,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4424192251703699,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 8676
+    },
+    {
+      "epoch": 0.44247021862647534,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 8677
+    },
+    {
+      "epoch": 0.4425212120825807,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 8678
+    },
+    {
+      "epoch": 0.4425722055386861,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 8679
+    },
+    {
+      "epoch": 0.4426231989947915,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4426741924508969,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 8681
+    },
+    {
+      "epoch": 0.44272518590700227,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 8682
+    },
+    {
+      "epoch": 0.44277617936310765,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 8683
+    },
+    {
+      "epoch": 0.4428271728192131,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 8684
+    },
+    {
+      "epoch": 0.44287816627531845,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4429291597314238,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 8686
+    },
+    {
+      "epoch": 0.44298015318752926,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 8687
+    },
+    {
+      "epoch": 0.44303114664363463,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 8688
+    },
+    {
+      "epoch": 0.44308214009974,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 8689
+    },
+    {
+      "epoch": 0.4431331335558454,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4431841270119508,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 8691
+    },
+    {
+      "epoch": 0.4432351204680562,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 8692
+    },
+    {
+      "epoch": 0.44328611392416156,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 8693
+    },
+    {
+      "epoch": 0.443337107380267,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 8694
+    },
+    {
+      "epoch": 0.44338810083637237,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 8695
+    },
+    {
+      "epoch": 0.44343909429247774,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 8696
+    },
+    {
+      "epoch": 0.4434900877485831,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 8697
+    },
+    {
+      "epoch": 0.44354108120468855,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 8698
+    },
+    {
+      "epoch": 0.4435920746607939,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 8699
+    },
+    {
+      "epoch": 0.4436430681168993,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4436940615730047,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4437450550291101,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4437960484852155,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 8703
+    },
+    {
+      "epoch": 0.44384704194132085,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 8704
+    },
+    {
+      "epoch": 0.4438980353974263,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 8705
+    },
+    {
+      "epoch": 0.44394902885353166,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 8706
+    },
+    {
+      "epoch": 0.44400002230963703,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 8707
+    },
+    {
+      "epoch": 0.44405101576574246,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 8708
+    },
+    {
+      "epoch": 0.44410200922184784,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4441530026779532,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4442039961340586,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 8711
+    },
+    {
+      "epoch": 0.444254989590164,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 8712
+    },
+    {
+      "epoch": 0.4443059830462694,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 8713
+    },
+    {
+      "epoch": 0.44435697650237477,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 8714
+    },
+    {
+      "epoch": 0.4444079699584802,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 8715
+    },
+    {
+      "epoch": 0.4444589634145856,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 8716
+    },
+    {
+      "epoch": 0.44450995687069095,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 8717
+    },
+    {
+      "epoch": 0.4445609503267963,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8718
+    },
+    {
+      "epoch": 0.44461194378290175,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 8719
+    },
+    {
+      "epoch": 0.4446629372390071,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4447139306951125,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 8721
+    },
+    {
+      "epoch": 0.44476492415121793,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 8722
+    },
+    {
+      "epoch": 0.4448159176073233,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4448669110634287,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8468,
+      "step": 8724
+    },
+    {
+      "epoch": 0.44491790451953406,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4449688979756395,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.835,
+      "step": 8726
+    },
+    {
+      "epoch": 0.44501989143174486,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 8727
+    },
+    {
+      "epoch": 0.44507088488785024,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 8728
+    },
+    {
+      "epoch": 0.44512187834395567,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 8729
+    },
+    {
+      "epoch": 0.44517287180006104,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 8730
+    },
+    {
+      "epoch": 0.4452238652561664,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 8731
+    },
+    {
+      "epoch": 0.4452748587122718,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 8732
+    },
+    {
+      "epoch": 0.4453258521683772,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 8733
+    },
+    {
+      "epoch": 0.4453768456244826,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 8734
+    },
+    {
+      "epoch": 0.445427839080588,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8735
+    },
+    {
+      "epoch": 0.4454788325366934,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8736
+    },
+    {
+      "epoch": 0.4455298259927988,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8737
+    },
+    {
+      "epoch": 0.44558081944890415,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 8738
+    },
+    {
+      "epoch": 0.44563181290500953,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 8739
+    },
+    {
+      "epoch": 0.44568280636111496,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44573379981722033,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4457847932733257,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 8742
+    },
+    {
+      "epoch": 0.44583578672943114,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 8743
+    },
+    {
+      "epoch": 0.4458867801855365,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4459377736416419,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 8745
+    },
+    {
+      "epoch": 0.44598876709774726,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4460397605538527,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 8747
+    },
+    {
+      "epoch": 0.44609075400995807,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 8748
+    },
+    {
+      "epoch": 0.44614174746606344,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 8749
+    },
+    {
+      "epoch": 0.4461927409221689,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 8750
+    },
+    {
+      "epoch": 0.44624373437827425,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4462947278343796,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 8752
+    },
+    {
+      "epoch": 0.446345721290485,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 8753
+    },
+    {
+      "epoch": 0.44639671474659043,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4464477082026958,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 8755
+    },
+    {
+      "epoch": 0.4464987016588012,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 8756
+    },
+    {
+      "epoch": 0.4465496951149066,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 8757
+    },
+    {
+      "epoch": 0.446600688571012,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 8758
+    },
+    {
+      "epoch": 0.44665168202711736,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 8759
+    },
+    {
+      "epoch": 0.44670267548322273,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 8760
+    },
+    {
+      "epoch": 0.44675366893932816,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 8761
+    },
+    {
+      "epoch": 0.44680466239543354,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 8762
+    },
+    {
+      "epoch": 0.4468556558515389,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 8763
+    },
+    {
+      "epoch": 0.44690664930764434,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 8764
+    },
+    {
+      "epoch": 0.4469576427637497,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 8765
+    },
+    {
+      "epoch": 0.4470086362198551,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 8766
+    },
+    {
+      "epoch": 0.44705962967596047,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4471106231320659,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 8768
+    },
+    {
+      "epoch": 0.4471616165881713,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 8769
+    },
+    {
+      "epoch": 0.44721261004427665,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4472636035003821,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 8771
+    },
+    {
+      "epoch": 0.44731459695648745,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7742,
+      "step": 8772
+    },
+    {
+      "epoch": 0.44736559041259283,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 8773
+    },
+    {
+      "epoch": 0.4474165838686982,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 8774
+    },
+    {
+      "epoch": 0.44746757732480363,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 8775
+    },
+    {
+      "epoch": 0.447518570780909,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 8776
+    },
+    {
+      "epoch": 0.4475695642370144,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 8777
+    },
+    {
+      "epoch": 0.4476205576931198,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4476715511492252,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 8779
+    },
+    {
+      "epoch": 0.44772254460533056,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 8780
+    },
+    {
+      "epoch": 0.44777353806143594,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 8781
+    },
+    {
+      "epoch": 0.44782453151754137,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 8782
+    },
+    {
+      "epoch": 0.44787552497364674,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 8783
+    },
+    {
+      "epoch": 0.4479265184297521,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 8784
+    },
+    {
+      "epoch": 0.44797751188585755,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4480285053419629,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 8786
+    },
+    {
+      "epoch": 0.4480794987980683,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 8787
+    },
+    {
+      "epoch": 0.4481304922541737,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 8788
+    },
+    {
+      "epoch": 0.4481814857102791,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 8789
+    },
+    {
+      "epoch": 0.4482324791663845,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7601,
+      "step": 8790
+    },
+    {
+      "epoch": 0.44828347262248985,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 8791
+    },
+    {
+      "epoch": 0.4483344660785953,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8792
+    },
+    {
+      "epoch": 0.44838545953470066,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 8793
+    },
+    {
+      "epoch": 0.44843645299080603,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 8794
+    },
+    {
+      "epoch": 0.4484874464469114,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 8795
+    },
+    {
+      "epoch": 0.44853843990301684,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 8796
+    },
+    {
+      "epoch": 0.4485894333591222,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 8797
+    },
+    {
+      "epoch": 0.4486404268152276,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 8798
+    },
+    {
+      "epoch": 0.448691420271333,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 8799
+    },
+    {
+      "epoch": 0.4487424137274384,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 8800
+    },
+    {
+      "epoch": 0.44879340718354377,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 8801
+    },
+    {
+      "epoch": 0.44884440063964914,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 8802
+    },
+    {
+      "epoch": 0.4488953940957546,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 8803
+    },
+    {
+      "epoch": 0.44894638755185995,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4489973810079653,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 8805
+    },
+    {
+      "epoch": 0.44904837446407075,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8806
+    },
+    {
+      "epoch": 0.44909936792017613,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4491503613762815,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 8808
+    },
+    {
+      "epoch": 0.4492013548323869,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 8809
+    },
+    {
+      "epoch": 0.4492523482884923,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4493033417445977,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 8811
+    },
+    {
+      "epoch": 0.44935433520070306,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4494053286568085,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 8813
+    },
+    {
+      "epoch": 0.44945632211291386,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 8814
+    },
+    {
+      "epoch": 0.44950731556901924,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 8815
+    },
+    {
+      "epoch": 0.4495583090251246,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 8816
+    },
+    {
+      "epoch": 0.44960930248123004,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 8817
+    },
+    {
+      "epoch": 0.4496602959373354,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8575,
+      "step": 8818
+    },
+    {
+      "epoch": 0.4497112893934408,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8399,
+      "step": 8819
+    },
+    {
+      "epoch": 0.4497622828495462,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4498132763056516,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 8821
+    },
+    {
+      "epoch": 0.449864269761757,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 8822
+    },
+    {
+      "epoch": 0.44991526321786235,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 8823
+    },
+    {
+      "epoch": 0.4499662566739678,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 8824
+    },
+    {
+      "epoch": 0.45001725013007315,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 8825
+    },
+    {
+      "epoch": 0.45006824358617853,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 8826
+    },
+    {
+      "epoch": 0.45011923704228396,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 8827
+    },
+    {
+      "epoch": 0.45017023049838933,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 8828
+    },
+    {
+      "epoch": 0.4502212239544947,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 8829
+    },
+    {
+      "epoch": 0.4502722174106001,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4503232108667055,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 8831
+    },
+    {
+      "epoch": 0.4503742043228109,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8518,
+      "step": 8832
+    },
+    {
+      "epoch": 0.45042519777891626,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 8833
+    },
+    {
+      "epoch": 0.4504761912350217,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 8834
+    },
+    {
+      "epoch": 0.45052718469112707,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 8835
+    },
+    {
+      "epoch": 0.45057817814723244,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 8836
+    },
+    {
+      "epoch": 0.4506291716033378,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 8837
+    },
+    {
+      "epoch": 0.45068016505944325,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 8838
+    },
+    {
+      "epoch": 0.4507311585155486,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 8839
+    },
+    {
+      "epoch": 0.450782151971654,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 8840
+    },
+    {
+      "epoch": 0.45083314542775943,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 8841
+    },
+    {
+      "epoch": 0.4508841388838648,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 8842
+    },
+    {
+      "epoch": 0.4509351323399702,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 8843
+    },
+    {
+      "epoch": 0.45098612579607555,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 8844
+    },
+    {
+      "epoch": 0.451037119252181,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 8845
+    },
+    {
+      "epoch": 0.45108811270828636,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 8846
+    },
+    {
+      "epoch": 0.45113910616439173,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 8847
+    },
+    {
+      "epoch": 0.45119009962049716,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 8848
+    },
+    {
+      "epoch": 0.45124109307660254,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 8849
+    },
+    {
+      "epoch": 0.4512920865327079,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4513430799888133,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 8851
+    },
+    {
+      "epoch": 0.4513940734449187,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 8852
+    },
+    {
+      "epoch": 0.4514450669010241,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 8853
+    },
+    {
+      "epoch": 0.45149606035712947,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 8854
+    },
+    {
+      "epoch": 0.4515470538132349,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4515980472693403,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 8856
+    },
+    {
+      "epoch": 0.45164904072544565,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 8857
+    },
+    {
+      "epoch": 0.451700034181551,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 8858
+    },
+    {
+      "epoch": 0.45175102763765645,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 8859
+    },
+    {
+      "epoch": 0.45180202109376183,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4518530145498672,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 8861
+    },
+    {
+      "epoch": 0.45190400800597263,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 8862
+    },
+    {
+      "epoch": 0.451955001462078,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 8863
+    },
+    {
+      "epoch": 0.4520059949181834,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 8864
+    },
+    {
+      "epoch": 0.45205698837428876,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 8865
+    },
+    {
+      "epoch": 0.4521079818303942,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 8866
+    },
+    {
+      "epoch": 0.45215897528649956,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 8867
+    },
+    {
+      "epoch": 0.45220996874260494,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 8868
+    },
+    {
+      "epoch": 0.45226096219871037,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 8869
+    },
+    {
+      "epoch": 0.45231195565481574,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4523629491109211,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 8871
+    },
+    {
+      "epoch": 0.4524139425670265,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 8872
+    },
+    {
+      "epoch": 0.4524649360231319,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 8873
+    },
+    {
+      "epoch": 0.4525159294792373,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 8874
+    },
+    {
+      "epoch": 0.4525669229353427,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 8875
+    },
+    {
+      "epoch": 0.4526179163914481,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 8876
+    },
+    {
+      "epoch": 0.4526689098475535,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 8877
+    },
+    {
+      "epoch": 0.45271990330365885,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 8878
+    },
+    {
+      "epoch": 0.45277089675976423,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 8879
+    },
+    {
+      "epoch": 0.45282189021586966,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45287288367197504,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 8881
+    },
+    {
+      "epoch": 0.4529238771280804,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 8882
+    },
+    {
+      "epoch": 0.45297487058418584,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 8883
+    },
+    {
+      "epoch": 0.4530258640402912,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 8884
+    },
+    {
+      "epoch": 0.4530768574963966,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 8885
+    },
+    {
+      "epoch": 0.45312785095250196,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 8886
+    },
+    {
+      "epoch": 0.4531788444086074,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 8887
+    },
+    {
+      "epoch": 0.45322983786471277,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 8888
+    },
+    {
+      "epoch": 0.45328083132081814,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 8889
+    },
+    {
+      "epoch": 0.4533318247769236,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45338281823302895,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 8891
+    },
+    {
+      "epoch": 0.4534338116891343,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 8892
+    },
+    {
+      "epoch": 0.4534848051452397,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 8893
+    },
+    {
+      "epoch": 0.45353579860134513,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 8894
+    },
+    {
+      "epoch": 0.4535867920574505,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4536377855135559,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 8896
+    },
+    {
+      "epoch": 0.4536887789696613,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 8897
+    },
+    {
+      "epoch": 0.4537397724257667,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 8898
+    },
+    {
+      "epoch": 0.45379076588187206,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 8899
+    },
+    {
+      "epoch": 0.45384175933797744,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 8900
+    },
+    {
+      "epoch": 0.45389275279408287,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 8901
+    },
+    {
+      "epoch": 0.45394374625018824,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 8902
+    },
+    {
+      "epoch": 0.4539947397062936,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 8903
+    },
+    {
+      "epoch": 0.45404573316239905,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8904
+    },
+    {
+      "epoch": 0.4540967266185044,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 8905
+    },
+    {
+      "epoch": 0.4541477200746098,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 8906
+    },
+    {
+      "epoch": 0.45419871353071517,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 8907
+    },
+    {
+      "epoch": 0.4542497069868206,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 8908
+    },
+    {
+      "epoch": 0.454300700442926,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 8909
+    },
+    {
+      "epoch": 0.45435169389903135,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 8910
+    },
+    {
+      "epoch": 0.4544026873551368,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 8911
+    },
+    {
+      "epoch": 0.45445368081124216,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 8912
+    },
+    {
+      "epoch": 0.45450467426734753,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 8913
+    },
+    {
+      "epoch": 0.4545556677234529,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 8914
+    },
+    {
+      "epoch": 0.45460666117955834,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 8915
+    },
+    {
+      "epoch": 0.4546576546356637,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 8916
+    },
+    {
+      "epoch": 0.4547086480917691,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 8917
+    },
+    {
+      "epoch": 0.4547596415478745,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 8918
+    },
+    {
+      "epoch": 0.4548106350039799,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 8919
+    },
+    {
+      "epoch": 0.45486162846008527,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 8920
+    },
+    {
+      "epoch": 0.45491262191619064,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 8921
+    },
+    {
+      "epoch": 0.45496361537229607,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 8922
+    },
+    {
+      "epoch": 0.45501460882840145,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 8923
+    },
+    {
+      "epoch": 0.4550656022845068,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 8924
+    },
+    {
+      "epoch": 0.45511659574061225,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 8925
+    },
+    {
+      "epoch": 0.4551675891967176,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 8926
+    },
+    {
+      "epoch": 0.455218582652823,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 8927
+    },
+    {
+      "epoch": 0.4552695761089284,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 8928
+    },
+    {
+      "epoch": 0.4553205695650338,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 8929
+    },
+    {
+      "epoch": 0.4553715630211392,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 8930
+    },
+    {
+      "epoch": 0.45542255647724456,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 8931
+    },
+    {
+      "epoch": 0.45547354993335,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 8932
+    },
+    {
+      "epoch": 0.45552454338945536,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 8933
+    },
+    {
+      "epoch": 0.45557553684556074,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 8934
+    },
+    {
+      "epoch": 0.4556265303016661,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 8935
+    },
+    {
+      "epoch": 0.45567752375777154,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 8936
+    },
+    {
+      "epoch": 0.4557285172138769,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 8937
+    },
+    {
+      "epoch": 0.4557795106699823,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 8938
+    },
+    {
+      "epoch": 0.4558305041260877,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 8939
+    },
+    {
+      "epoch": 0.4558814975821931,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45593249103829847,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 8941
+    },
+    {
+      "epoch": 0.45598348449440385,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 8942
+    },
+    {
+      "epoch": 0.4560344779505093,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 8943
+    },
+    {
+      "epoch": 0.45608547140661465,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 8944
+    },
+    {
+      "epoch": 0.45613646486272,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 8945
+    },
+    {
+      "epoch": 0.45618745831882546,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 8946
+    },
+    {
+      "epoch": 0.45623845177493083,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 8947
+    },
+    {
+      "epoch": 0.4562894452310362,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 8948
+    },
+    {
+      "epoch": 0.4563404386871416,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 8949
+    },
+    {
+      "epoch": 0.456391432143247,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4564424255993524,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 8951
+    },
+    {
+      "epoch": 0.45649341905545776,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 8952
+    },
+    {
+      "epoch": 0.4565444125115632,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 8953
+    },
+    {
+      "epoch": 0.45659540596766857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 8954
+    },
+    {
+      "epoch": 0.45664639942377394,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4566973928798793,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 8956
+    },
+    {
+      "epoch": 0.45674838633598475,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 8957
+    },
+    {
+      "epoch": 0.4567993797920901,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 8958
+    },
+    {
+      "epoch": 0.4568503732481955,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 8959
+    },
+    {
+      "epoch": 0.4569013667043009,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4569523601604063,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 8961
+    },
+    {
+      "epoch": 0.4570033536165117,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 8962
+    },
+    {
+      "epoch": 0.45705434707261705,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 8963
+    },
+    {
+      "epoch": 0.4571053405287225,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8964
+    },
+    {
+      "epoch": 0.45715633398482786,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 8965
+    },
+    {
+      "epoch": 0.45720732744093323,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 8966
+    },
+    {
+      "epoch": 0.45725832089703866,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 8967
+    },
+    {
+      "epoch": 0.45730931435314404,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 8968
+    },
+    {
+      "epoch": 0.4573603078092494,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 8969
+    },
+    {
+      "epoch": 0.4574113012653548,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4574622947214602,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 8971
+    },
+    {
+      "epoch": 0.4575132881775656,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 8972
+    },
+    {
+      "epoch": 0.45756428163367097,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 8973
+    },
+    {
+      "epoch": 0.4576152750897764,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 8974
+    },
+    {
+      "epoch": 0.45766626854588177,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 8975
+    },
+    {
+      "epoch": 0.45771726200198715,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 8976
+    },
+    {
+      "epoch": 0.4577682554580925,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 8977
+    },
+    {
+      "epoch": 0.45781924891419795,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 8978
+    },
+    {
+      "epoch": 0.4578702423703033,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 8979
+    },
+    {
+      "epoch": 0.4579212358264087,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 8980
+    },
+    {
+      "epoch": 0.45797222928251413,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 8981
+    },
+    {
+      "epoch": 0.4580232227386195,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 8982
+    },
+    {
+      "epoch": 0.4580742161947249,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 8983
+    },
+    {
+      "epoch": 0.45812520965083026,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 8984
+    },
+    {
+      "epoch": 0.4581762031069357,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 8985
+    },
+    {
+      "epoch": 0.45822719656304106,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 8986
+    },
+    {
+      "epoch": 0.45827819001914644,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 8987
+    },
+    {
+      "epoch": 0.45832918347525187,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 8988
+    },
+    {
+      "epoch": 0.45838017693135724,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7805,
+      "step": 8989
+    },
+    {
+      "epoch": 0.4584311703874626,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 8990
+    },
+    {
+      "epoch": 0.458482163843568,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 8991
+    },
+    {
+      "epoch": 0.4585331572996734,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 8992
+    },
+    {
+      "epoch": 0.4585841507557788,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 8993
+    },
+    {
+      "epoch": 0.45863514421188417,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 8994
+    },
+    {
+      "epoch": 0.4586861376679896,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 8995
+    },
+    {
+      "epoch": 0.458737131124095,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 8996
+    },
+    {
+      "epoch": 0.45878812458020035,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 8997
+    },
+    {
+      "epoch": 0.4588391180363057,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 8998
+    },
+    {
+      "epoch": 0.45889011149241116,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 8999
+    },
+    {
+      "epoch": 0.45894110494851653,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4589920984046219,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 9001
+    },
+    {
+      "epoch": 0.45904309186072734,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 9002
+    },
+    {
+      "epoch": 0.4590940853168327,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 9003
+    },
+    {
+      "epoch": 0.4591450787729381,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 9004
+    },
+    {
+      "epoch": 0.45919607222904346,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 9005
+    },
+    {
+      "epoch": 0.4592470656851489,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9006
+    },
+    {
+      "epoch": 0.45929805914125427,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 9007
+    },
+    {
+      "epoch": 0.45934905259735964,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 9008
+    },
+    {
+      "epoch": 0.45940004605346507,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 9009
+    },
+    {
+      "epoch": 0.45945103950957045,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 9010
+    },
+    {
+      "epoch": 0.4595020329656758,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9011
+    },
+    {
+      "epoch": 0.4595530264217812,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 9012
+    },
+    {
+      "epoch": 0.4596040198778866,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 9013
+    },
+    {
+      "epoch": 0.459655013333992,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 9014
+    },
+    {
+      "epoch": 0.4597060067900974,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 9015
+    },
+    {
+      "epoch": 0.4597570002462028,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 9016
+    },
+    {
+      "epoch": 0.4598079937023082,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 9017
+    },
+    {
+      "epoch": 0.45985898715841356,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 9018
+    },
+    {
+      "epoch": 0.45990998061451893,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 9019
+    },
+    {
+      "epoch": 0.45996097407062436,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46001196752672974,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 9021
+    },
+    {
+      "epoch": 0.4600629609828351,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 9022
+    },
+    {
+      "epoch": 0.46011395443894054,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 9023
+    },
+    {
+      "epoch": 0.4601649478950459,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 9024
+    },
+    {
+      "epoch": 0.4602159413511513,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 9025
+    },
+    {
+      "epoch": 0.46026693480725667,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7879,
+      "step": 9026
+    },
+    {
+      "epoch": 0.4603179282633621,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 9027
+    },
+    {
+      "epoch": 0.46036892171946747,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 9028
+    },
+    {
+      "epoch": 0.46041991517557285,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 9029
+    },
+    {
+      "epoch": 0.4604709086316783,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8401,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46052190208778365,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 9031
+    },
+    {
+      "epoch": 0.460572895543889,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 9032
+    },
+    {
+      "epoch": 0.4606238889999944,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 9033
+    },
+    {
+      "epoch": 0.46067488245609983,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 9034
+    },
+    {
+      "epoch": 0.4607258759122052,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 9035
+    },
+    {
+      "epoch": 0.4607768693683106,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 9036
+    },
+    {
+      "epoch": 0.460827862824416,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 9037
+    },
+    {
+      "epoch": 0.4608788562805214,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 9038
+    },
+    {
+      "epoch": 0.46092984973662676,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8561,
+      "step": 9039
+    },
+    {
+      "epoch": 0.46098084319273214,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 9040
+    },
+    {
+      "epoch": 0.46103183664883757,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 9041
+    },
+    {
+      "epoch": 0.46108283010494294,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 9042
+    },
+    {
+      "epoch": 0.4611338235610483,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 9043
+    },
+    {
+      "epoch": 0.46118481701715375,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 9044
+    },
+    {
+      "epoch": 0.4612358104732591,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 9045
+    },
+    {
+      "epoch": 0.4612868039293645,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 9046
+    },
+    {
+      "epoch": 0.46133779738546987,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9047
+    },
+    {
+      "epoch": 0.4613887908415753,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 9048
+    },
+    {
+      "epoch": 0.4614397842976807,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 9049
+    },
+    {
+      "epoch": 0.46149077775378605,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4615417712098915,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8377,
+      "step": 9051
+    },
+    {
+      "epoch": 0.46159276466599686,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 9052
+    },
+    {
+      "epoch": 0.46164375812210223,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 9053
+    },
+    {
+      "epoch": 0.4616947515782076,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 9054
+    },
+    {
+      "epoch": 0.46174574503431304,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 9055
+    },
+    {
+      "epoch": 0.4617967384904184,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 9056
+    },
+    {
+      "epoch": 0.4618477319465238,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 9057
+    },
+    {
+      "epoch": 0.4618987254026292,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 9058
+    },
+    {
+      "epoch": 0.4619497188587346,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 9059
+    },
+    {
+      "epoch": 0.46200071231483997,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 9060
+    },
+    {
+      "epoch": 0.46205170577094534,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 9061
+    },
+    {
+      "epoch": 0.4621026992270508,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 9062
+    },
+    {
+      "epoch": 0.46215369268315615,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 9063
+    },
+    {
+      "epoch": 0.4622046861392615,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 9064
+    },
+    {
+      "epoch": 0.46225567959536695,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 9065
+    },
+    {
+      "epoch": 0.4623066730514723,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 9066
+    },
+    {
+      "epoch": 0.4623576665075777,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8427,
+      "step": 9067
+    },
+    {
+      "epoch": 0.4624086599636831,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 9068
+    },
+    {
+      "epoch": 0.4624596534197885,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 9069
+    },
+    {
+      "epoch": 0.4625106468758939,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8448,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46256164033199926,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 9071
+    },
+    {
+      "epoch": 0.4626126337881047,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 9072
+    },
+    {
+      "epoch": 0.46266362724421006,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 9073
+    },
+    {
+      "epoch": 0.46271462070031544,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 9074
+    },
+    {
+      "epoch": 0.4627656141564208,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 9075
+    },
+    {
+      "epoch": 0.46281660761252624,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 9076
+    },
+    {
+      "epoch": 0.4628676010686316,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 9077
+    },
+    {
+      "epoch": 0.462918594524737,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8453,
+      "step": 9078
+    },
+    {
+      "epoch": 0.4629695879808424,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 9079
+    },
+    {
+      "epoch": 0.4630205814369478,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4630715748930532,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 9081
+    },
+    {
+      "epoch": 0.46312256834915855,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 9082
+    },
+    {
+      "epoch": 0.463173561805264,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 9083
+    },
+    {
+      "epoch": 0.46322455526136935,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 9084
+    },
+    {
+      "epoch": 0.46327554871747473,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 9085
+    },
+    {
+      "epoch": 0.46332654217358016,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 9086
+    },
+    {
+      "epoch": 0.46337753562968553,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 9087
+    },
+    {
+      "epoch": 0.4634285290857909,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8351,
+      "step": 9088
+    },
+    {
+      "epoch": 0.4634795225418963,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 9089
+    },
+    {
+      "epoch": 0.4635305159980017,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 9090
+    },
+    {
+      "epoch": 0.4635815094541071,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 9091
+    },
+    {
+      "epoch": 0.46363250291021246,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 9092
+    },
+    {
+      "epoch": 0.4636834963663179,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 9093
+    },
+    {
+      "epoch": 0.46373448982242327,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 9094
+    },
+    {
+      "epoch": 0.46378548327852864,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 9095
+    },
+    {
+      "epoch": 0.463836476734634,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 9096
+    },
+    {
+      "epoch": 0.46388747019073945,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 9097
+    },
+    {
+      "epoch": 0.4639384636468448,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 9098
+    },
+    {
+      "epoch": 0.4639894571029502,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 9099
+    },
+    {
+      "epoch": 0.46404045055905563,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 9100
+    },
+    {
+      "epoch": 0.464091444015161,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 9101
+    },
+    {
+      "epoch": 0.4641424374712664,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9102
+    },
+    {
+      "epoch": 0.46419343092737175,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 9103
+    },
+    {
+      "epoch": 0.4642444243834772,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 9104
+    },
+    {
+      "epoch": 0.46429541783958256,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 9105
+    },
+    {
+      "epoch": 0.46434641129568793,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 9106
+    },
+    {
+      "epoch": 0.46439740475179336,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 9107
+    },
+    {
+      "epoch": 0.46444839820789874,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 9108
+    },
+    {
+      "epoch": 0.4644993916640041,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 9109
+    },
+    {
+      "epoch": 0.4645503851201095,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4646013785762149,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 9111
+    },
+    {
+      "epoch": 0.4646523720323203,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 9112
+    },
+    {
+      "epoch": 0.46470336548842567,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 9113
+    },
+    {
+      "epoch": 0.4647543589445311,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 9114
+    },
+    {
+      "epoch": 0.4648053524006365,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 9115
+    },
+    {
+      "epoch": 0.46485634585674185,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 9116
+    },
+    {
+      "epoch": 0.4649073393128472,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 9117
+    },
+    {
+      "epoch": 0.46495833276895265,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 9118
+    },
+    {
+      "epoch": 0.46500932622505803,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 9119
+    },
+    {
+      "epoch": 0.4650603196811634,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9120
+    },
+    {
+      "epoch": 0.46511131313726883,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 9121
+    },
+    {
+      "epoch": 0.4651623065933742,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 9122
+    },
+    {
+      "epoch": 0.4652133000494796,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 9123
+    },
+    {
+      "epoch": 0.46526429350558496,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 9124
+    },
+    {
+      "epoch": 0.4653152869616904,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 9125
+    },
+    {
+      "epoch": 0.46536628041779576,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 9126
+    },
+    {
+      "epoch": 0.46541727387390114,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 9127
+    },
+    {
+      "epoch": 0.46546826733000657,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 9128
+    },
+    {
+      "epoch": 0.46551926078611194,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 9129
+    },
+    {
+      "epoch": 0.4655702542422173,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4656212476983227,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 9131
+    },
+    {
+      "epoch": 0.4656722411544281,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8476,
+      "step": 9132
+    },
+    {
+      "epoch": 0.4657232346105335,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 9133
+    },
+    {
+      "epoch": 0.4657742280666389,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 9134
+    },
+    {
+      "epoch": 0.4658252215227443,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8544,
+      "step": 9135
+    },
+    {
+      "epoch": 0.4658762149788497,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 9136
+    },
+    {
+      "epoch": 0.46592720843495505,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9137
+    },
+    {
+      "epoch": 0.46597820189106043,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 9138
+    },
+    {
+      "epoch": 0.46602919534716586,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 9139
+    },
+    {
+      "epoch": 0.46608018880327123,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4661311822593766,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 9141
+    },
+    {
+      "epoch": 0.46618217571548204,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 9142
+    },
+    {
+      "epoch": 0.4662331691715874,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 9143
+    },
+    {
+      "epoch": 0.4662841626276928,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 9144
+    },
+    {
+      "epoch": 0.46633515608379816,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 9145
+    },
+    {
+      "epoch": 0.4663861495399036,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9146
+    },
+    {
+      "epoch": 0.46643714299600897,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 9147
+    },
+    {
+      "epoch": 0.46648813645211434,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 9148
+    },
+    {
+      "epoch": 0.4665391299082198,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 9149
+    },
+    {
+      "epoch": 0.46659012336432515,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4666411168204305,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 9151
+    },
+    {
+      "epoch": 0.4666921102765359,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 9152
+    },
+    {
+      "epoch": 0.46674310373264133,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 9153
+    },
+    {
+      "epoch": 0.4667940971887467,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.862,
+      "step": 9154
+    },
+    {
+      "epoch": 0.4668450906448521,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8378,
+      "step": 9155
+    },
+    {
+      "epoch": 0.4668960841009575,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7757,
+      "step": 9156
+    },
+    {
+      "epoch": 0.4669470775570629,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 9157
+    },
+    {
+      "epoch": 0.46699807101316826,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7665,
+      "step": 9158
+    },
+    {
+      "epoch": 0.46704906446927363,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 9159
+    },
+    {
+      "epoch": 0.46710005792537906,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 9160
+    },
+    {
+      "epoch": 0.46715105138148444,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9161
+    },
+    {
+      "epoch": 0.4672020448375898,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 9162
+    },
+    {
+      "epoch": 0.46725303829369524,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 9163
+    },
+    {
+      "epoch": 0.4673040317498006,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 9164
+    },
+    {
+      "epoch": 0.467355025205906,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8595,
+      "step": 9165
+    },
+    {
+      "epoch": 0.46740601866201137,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 9166
+    },
+    {
+      "epoch": 0.4674570121181168,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 9167
+    },
+    {
+      "epoch": 0.4675080055742222,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 9168
+    },
+    {
+      "epoch": 0.46755899903032755,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 9169
+    },
+    {
+      "epoch": 0.467609992486433,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 9170
+    },
+    {
+      "epoch": 0.46766098594253835,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 9171
+    },
+    {
+      "epoch": 0.46771197939864373,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 9172
+    },
+    {
+      "epoch": 0.4677629728547491,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8434,
+      "step": 9173
+    },
+    {
+      "epoch": 0.46781396631085453,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 9174
+    },
+    {
+      "epoch": 0.4678649597669599,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 9175
+    },
+    {
+      "epoch": 0.4679159532230653,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 9176
+    },
+    {
+      "epoch": 0.4679669466791707,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 9177
+    },
+    {
+      "epoch": 0.4680179401352761,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 9178
+    },
+    {
+      "epoch": 0.46806893359138146,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 9179
+    },
+    {
+      "epoch": 0.46811992704748684,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 9180
+    },
+    {
+      "epoch": 0.46817092050359227,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 9181
+    },
+    {
+      "epoch": 0.46822191395969764,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 9182
+    },
+    {
+      "epoch": 0.468272907415803,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 9183
+    },
+    {
+      "epoch": 0.46832390087190845,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 9184
+    },
+    {
+      "epoch": 0.4683748943280138,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 9185
+    },
+    {
+      "epoch": 0.4684258877841192,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8377,
+      "step": 9186
+    },
+    {
+      "epoch": 0.4684768812402246,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 9187
+    },
+    {
+      "epoch": 0.46852787469633,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 9188
+    },
+    {
+      "epoch": 0.4685788681524354,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 9189
+    },
+    {
+      "epoch": 0.46862986160854075,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4686808550646462,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 9191
+    },
+    {
+      "epoch": 0.46873184852075156,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8475,
+      "step": 9192
+    },
+    {
+      "epoch": 0.46878284197685693,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 9193
+    },
+    {
+      "epoch": 0.4688338354329623,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 9194
+    },
+    {
+      "epoch": 0.46888482888906774,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 9195
+    },
+    {
+      "epoch": 0.4689358223451731,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 9196
+    },
+    {
+      "epoch": 0.4689868158012785,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 9197
+    },
+    {
+      "epoch": 0.4690378092573839,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 9198
+    },
+    {
+      "epoch": 0.4690888027134893,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 9199
+    },
+    {
+      "epoch": 0.46913979616959467,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 9200
+    },
+    {
+      "epoch": 0.46919078962570004,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 9201
+    },
+    {
+      "epoch": 0.4692417830818055,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9202
+    },
+    {
+      "epoch": 0.46929277653791085,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 9203
+    },
+    {
+      "epoch": 0.4693437699940162,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 9204
+    },
+    {
+      "epoch": 0.46939476345012165,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 9205
+    },
+    {
+      "epoch": 0.46944575690622703,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 9206
+    },
+    {
+      "epoch": 0.4694967503623324,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 9207
+    },
+    {
+      "epoch": 0.4695477438184378,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 9208
+    },
+    {
+      "epoch": 0.4695987372745432,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 9209
+    },
+    {
+      "epoch": 0.4696497307306486,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 9210
+    },
+    {
+      "epoch": 0.46970072418675396,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 9211
+    },
+    {
+      "epoch": 0.4697517176428594,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 9212
+    },
+    {
+      "epoch": 0.46980271109896476,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 9213
+    },
+    {
+      "epoch": 0.46985370455507014,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.764,
+      "step": 9214
+    },
+    {
+      "epoch": 0.4699046980111755,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 9215
+    },
+    {
+      "epoch": 0.46995569146728094,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 9216
+    },
+    {
+      "epoch": 0.4700066849233863,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 9217
+    },
+    {
+      "epoch": 0.4700576783794917,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 9218
+    },
+    {
+      "epoch": 0.4701086718355971,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 9219
+    },
+    {
+      "epoch": 0.4701596652917025,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 9220
+    },
+    {
+      "epoch": 0.4702106587478079,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 9221
+    },
+    {
+      "epoch": 0.47026165220391325,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 9222
+    },
+    {
+      "epoch": 0.4703126456600187,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 9223
+    },
+    {
+      "epoch": 0.47036363911612405,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 9224
+    },
+    {
+      "epoch": 0.47041463257222943,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 9225
+    },
+    {
+      "epoch": 0.47046562602833486,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 9226
+    },
+    {
+      "epoch": 0.47051661948444023,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 9227
+    },
+    {
+      "epoch": 0.4705676129405456,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 9228
+    },
+    {
+      "epoch": 0.470618606396651,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 9229
+    },
+    {
+      "epoch": 0.4706695998527564,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 9230
+    },
+    {
+      "epoch": 0.4707205933088618,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 9231
+    },
+    {
+      "epoch": 0.47077158676496716,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.865,
+      "step": 9232
+    },
+    {
+      "epoch": 0.4708225802210726,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 9233
+    },
+    {
+      "epoch": 0.47087357367717797,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 9234
+    },
+    {
+      "epoch": 0.47092456713328334,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 9235
+    },
+    {
+      "epoch": 0.4709755605893887,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 9236
+    },
+    {
+      "epoch": 0.47102655404549415,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 9237
+    },
+    {
+      "epoch": 0.4710775475015995,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 9238
+    },
+    {
+      "epoch": 0.4711285409577049,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 9239
+    },
+    {
+      "epoch": 0.47117953441381033,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4712305278699157,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 9241
+    },
+    {
+      "epoch": 0.4712815213260211,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 9242
+    },
+    {
+      "epoch": 0.47133251478212645,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 9243
+    },
+    {
+      "epoch": 0.4713835082382319,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8298,
+      "step": 9244
+    },
+    {
+      "epoch": 0.47143450169433726,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 9245
+    },
+    {
+      "epoch": 0.47148549515044264,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 9246
+    },
+    {
+      "epoch": 0.47153648860654807,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 9247
+    },
+    {
+      "epoch": 0.47158748206265344,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 9248
+    },
+    {
+      "epoch": 0.4716384755187588,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 9249
+    },
+    {
+      "epoch": 0.4716894689748642,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4717404624309696,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 9251
+    },
+    {
+      "epoch": 0.471791455887075,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 9252
+    },
+    {
+      "epoch": 0.47184244934318037,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 9253
+    },
+    {
+      "epoch": 0.4718934427992858,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8464,
+      "step": 9254
+    },
+    {
+      "epoch": 0.4719444362553912,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 9255
+    },
+    {
+      "epoch": 0.47199542971149655,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 9256
+    },
+    {
+      "epoch": 0.4720464231676019,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 9257
+    },
+    {
+      "epoch": 0.47209741662370736,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 9258
+    },
+    {
+      "epoch": 0.47214841007981273,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 9259
+    },
+    {
+      "epoch": 0.4721994035359181,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47225039699202354,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7766,
+      "step": 9261
+    },
+    {
+      "epoch": 0.4723013904481289,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 9262
+    },
+    {
+      "epoch": 0.4723523839042343,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 9263
+    },
+    {
+      "epoch": 0.47240337736033966,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 9264
+    },
+    {
+      "epoch": 0.4724543708164451,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9265
+    },
+    {
+      "epoch": 0.47250536427255047,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 9266
+    },
+    {
+      "epoch": 0.47255635772865584,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 9267
+    },
+    {
+      "epoch": 0.47260735118476127,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 9268
+    },
+    {
+      "epoch": 0.47265834464086665,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 9269
+    },
+    {
+      "epoch": 0.472709338096972,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4727603315530774,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 9271
+    },
+    {
+      "epoch": 0.4728113250091828,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 9272
+    },
+    {
+      "epoch": 0.4728623184652882,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 9273
+    },
+    {
+      "epoch": 0.4729133119213936,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 9274
+    },
+    {
+      "epoch": 0.472964305377499,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 9275
+    },
+    {
+      "epoch": 0.4730152988336044,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 9276
+    },
+    {
+      "epoch": 0.47306629228970976,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9277
+    },
+    {
+      "epoch": 0.47311728574581513,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 9278
+    },
+    {
+      "epoch": 0.47316827920192056,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 9279
+    },
+    {
+      "epoch": 0.47321927265802594,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 9280
+    },
+    {
+      "epoch": 0.4732702661141313,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 9281
+    },
+    {
+      "epoch": 0.47332125957023674,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 9282
+    },
+    {
+      "epoch": 0.4733722530263421,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 9283
+    },
+    {
+      "epoch": 0.4734232464824475,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 9284
+    },
+    {
+      "epoch": 0.47347423993855287,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 9285
+    },
+    {
+      "epoch": 0.4735252333946583,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 9286
+    },
+    {
+      "epoch": 0.47357622685076367,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 9287
+    },
+    {
+      "epoch": 0.47362722030686905,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 9288
+    },
+    {
+      "epoch": 0.4736782137629745,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 9289
+    },
+    {
+      "epoch": 0.47372920721907985,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 9290
+    },
+    {
+      "epoch": 0.4737802006751852,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 9291
+    },
+    {
+      "epoch": 0.4738311941312906,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 9292
+    },
+    {
+      "epoch": 0.47388218758739603,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 9293
+    },
+    {
+      "epoch": 0.4739331810435014,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 9294
+    },
+    {
+      "epoch": 0.4739841744996068,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 9295
+    },
+    {
+      "epoch": 0.4740351679557122,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 9296
+    },
+    {
+      "epoch": 0.4740861614118176,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 9297
+    },
+    {
+      "epoch": 0.47413715486792296,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 9298
+    },
+    {
+      "epoch": 0.47418814832402834,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 9299
+    },
+    {
+      "epoch": 0.47423914178013377,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 9300
+    },
+    {
+      "epoch": 0.47429013523623914,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 9301
+    },
+    {
+      "epoch": 0.4743411286923445,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 9302
+    },
+    {
+      "epoch": 0.47439212214844995,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 9303
+    },
+    {
+      "epoch": 0.4744431156045553,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 9304
+    },
+    {
+      "epoch": 0.4744941090606607,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 9305
+    },
+    {
+      "epoch": 0.47454510251676607,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 9306
+    },
+    {
+      "epoch": 0.4745960959728715,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 9307
+    },
+    {
+      "epoch": 0.4746470894289769,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 9308
+    },
+    {
+      "epoch": 0.47469808288508225,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 9309
+    },
+    {
+      "epoch": 0.4747490763411876,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 9310
+    },
+    {
+      "epoch": 0.47480006979729306,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 9311
+    },
+    {
+      "epoch": 0.47485106325339843,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 9312
+    },
+    {
+      "epoch": 0.4749020567095038,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 9313
+    },
+    {
+      "epoch": 0.47495305016560924,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 9314
+    },
+    {
+      "epoch": 0.4750040436217146,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 9315
+    },
+    {
+      "epoch": 0.47505503707782,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 9316
+    },
+    {
+      "epoch": 0.47510603053392536,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 9317
+    },
+    {
+      "epoch": 0.4751570239900308,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 9318
+    },
+    {
+      "epoch": 0.47520801744613617,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 9319
+    },
+    {
+      "epoch": 0.47525901090224154,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 9320
+    },
+    {
+      "epoch": 0.47531000435834697,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 9321
+    },
+    {
+      "epoch": 0.47536099781445235,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 9322
+    },
+    {
+      "epoch": 0.4754119912705577,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 9323
+    },
+    {
+      "epoch": 0.4754629847266631,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 9324
+    },
+    {
+      "epoch": 0.4755139781827685,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9325
+    },
+    {
+      "epoch": 0.4755649716388739,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 9326
+    },
+    {
+      "epoch": 0.4756159650949793,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 9327
+    },
+    {
+      "epoch": 0.4756669585510847,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 9328
+    },
+    {
+      "epoch": 0.4757179520071901,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 9329
+    },
+    {
+      "epoch": 0.47576894546329546,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9330
+    },
+    {
+      "epoch": 0.47581993891940083,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 9331
+    },
+    {
+      "epoch": 0.47587093237550626,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 9332
+    },
+    {
+      "epoch": 0.47592192583161164,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 9333
+    },
+    {
+      "epoch": 0.475972919287717,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 9334
+    },
+    {
+      "epoch": 0.47602391274382244,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 9335
+    },
+    {
+      "epoch": 0.4760749061999278,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 9336
+    },
+    {
+      "epoch": 0.4761258996560332,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 9337
+    },
+    {
+      "epoch": 0.47617689311213857,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 9338
+    },
+    {
+      "epoch": 0.476227886568244,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 9339
+    },
+    {
+      "epoch": 0.47627888002434937,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 9340
+    },
+    {
+      "epoch": 0.47632987348045475,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 9341
+    },
+    {
+      "epoch": 0.4763808669365602,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 9342
+    },
+    {
+      "epoch": 0.47643186039266555,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 9343
+    },
+    {
+      "epoch": 0.4764828538487709,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 9344
+    },
+    {
+      "epoch": 0.4765338473048763,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 9345
+    },
+    {
+      "epoch": 0.47658484076098173,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 9346
+    },
+    {
+      "epoch": 0.4766358342170871,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 9347
+    },
+    {
+      "epoch": 0.4766868276731925,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 9348
+    },
+    {
+      "epoch": 0.4767378211292979,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7802,
+      "step": 9349
+    },
+    {
+      "epoch": 0.4767888145854033,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 9350
+    },
+    {
+      "epoch": 0.47683980804150866,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 9351
+    },
+    {
+      "epoch": 0.47689080149761404,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 9352
+    },
+    {
+      "epoch": 0.47694179495371947,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 9353
+    },
+    {
+      "epoch": 0.47699278840982484,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 9354
+    },
+    {
+      "epoch": 0.4770437818659302,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 9355
+    },
+    {
+      "epoch": 0.47709477532203565,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 9356
+    },
+    {
+      "epoch": 0.477145768778141,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 9357
+    },
+    {
+      "epoch": 0.4771967622342464,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 9358
+    },
+    {
+      "epoch": 0.47724775569035177,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 9359
+    },
+    {
+      "epoch": 0.4772987491464572,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7808,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4773497426025626,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 9361
+    },
+    {
+      "epoch": 0.47740073605866795,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 9362
+    },
+    {
+      "epoch": 0.4774517295147734,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 9363
+    },
+    {
+      "epoch": 0.47750272297087876,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 9364
+    },
+    {
+      "epoch": 0.47755371642698413,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 9365
+    },
+    {
+      "epoch": 0.4776047098830895,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 9366
+    },
+    {
+      "epoch": 0.47765570333919494,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 9367
+    },
+    {
+      "epoch": 0.4777066967953003,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 9368
+    },
+    {
+      "epoch": 0.4777576902514057,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 9369
+    },
+    {
+      "epoch": 0.4778086837075111,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4778596771636165,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 9371
+    },
+    {
+      "epoch": 0.47791067061972187,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 9372
+    },
+    {
+      "epoch": 0.47796166407582724,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 9373
+    },
+    {
+      "epoch": 0.47801265753193267,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 9374
+    },
+    {
+      "epoch": 0.47806365098803805,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 9375
+    },
+    {
+      "epoch": 0.4781146444441434,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 9376
+    },
+    {
+      "epoch": 0.47816563790024885,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 9377
+    },
+    {
+      "epoch": 0.4782166313563542,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 9378
+    },
+    {
+      "epoch": 0.4782676248124596,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 9379
+    },
+    {
+      "epoch": 0.478318618268565,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 9380
+    },
+    {
+      "epoch": 0.4783696117246704,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 9381
+    },
+    {
+      "epoch": 0.4784206051807758,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 9382
+    },
+    {
+      "epoch": 0.47847159863688116,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 9383
+    },
+    {
+      "epoch": 0.4785225920929866,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 9384
+    },
+    {
+      "epoch": 0.47857358554909196,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 9385
+    },
+    {
+      "epoch": 0.47862457900519734,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 9386
+    },
+    {
+      "epoch": 0.4786755724613027,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 9387
+    },
+    {
+      "epoch": 0.47872656591740814,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 9388
+    },
+    {
+      "epoch": 0.4787775593735135,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 9389
+    },
+    {
+      "epoch": 0.4788285528296189,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4788795462857243,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 9391
+    },
+    {
+      "epoch": 0.4789305397418297,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 9392
+    },
+    {
+      "epoch": 0.47898153319793507,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 9393
+    },
+    {
+      "epoch": 0.47903252665404045,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 9394
+    },
+    {
+      "epoch": 0.4790835201101459,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 9395
+    },
+    {
+      "epoch": 0.47913451356625125,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9396
+    },
+    {
+      "epoch": 0.4791855070223566,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 9397
+    },
+    {
+      "epoch": 0.47923650047846206,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 9398
+    },
+    {
+      "epoch": 0.47928749393456743,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 9399
+    },
+    {
+      "epoch": 0.4793384873906728,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 9400
+    },
+    {
+      "epoch": 0.4793894808467782,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9401
+    },
+    {
+      "epoch": 0.4794404743028836,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 9402
+    },
+    {
+      "epoch": 0.479491467758989,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7606,
+      "step": 9403
+    },
+    {
+      "epoch": 0.47954246121509436,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 9404
+    },
+    {
+      "epoch": 0.4795934546711998,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 9405
+    },
+    {
+      "epoch": 0.47964444812730517,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 9406
+    },
+    {
+      "epoch": 0.47969544158341054,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 9407
+    },
+    {
+      "epoch": 0.4797464350395159,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 9408
+    },
+    {
+      "epoch": 0.47979742849562135,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 9409
+    },
+    {
+      "epoch": 0.4798484219517267,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 9410
+    },
+    {
+      "epoch": 0.4798994154078321,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 9411
+    },
+    {
+      "epoch": 0.4799504088639375,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 9412
+    },
+    {
+      "epoch": 0.4800014023200429,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7767,
+      "step": 9413
+    },
+    {
+      "epoch": 0.4800523957761483,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 9414
+    },
+    {
+      "epoch": 0.48010338923225365,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 9415
+    },
+    {
+      "epoch": 0.4801543826883591,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 9416
+    },
+    {
+      "epoch": 0.48020537614446446,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 9417
+    },
+    {
+      "epoch": 0.48025636960056983,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 9418
+    },
+    {
+      "epoch": 0.48030736305667526,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 9419
+    },
+    {
+      "epoch": 0.48035835651278064,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 9420
+    },
+    {
+      "epoch": 0.480409349968886,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 9421
+    },
+    {
+      "epoch": 0.4804603434249914,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 9422
+    },
+    {
+      "epoch": 0.4805113368810968,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 9423
+    },
+    {
+      "epoch": 0.4805623303372022,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9424
+    },
+    {
+      "epoch": 0.48061332379330757,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 9425
+    },
+    {
+      "epoch": 0.480664317249413,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8495,
+      "step": 9426
+    },
+    {
+      "epoch": 0.4807153107055184,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 9427
+    },
+    {
+      "epoch": 0.48076630416162375,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 9428
+    },
+    {
+      "epoch": 0.4808172976177291,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 9429
+    },
+    {
+      "epoch": 0.48086829107383455,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4809192845299399,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8513,
+      "step": 9431
+    },
+    {
+      "epoch": 0.4809702779860453,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 9432
+    },
+    {
+      "epoch": 0.48102127144215073,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 9433
+    },
+    {
+      "epoch": 0.4810722648982561,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 9434
+    },
+    {
+      "epoch": 0.4811232583543615,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 9435
+    },
+    {
+      "epoch": 0.48117425181046686,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 9436
+    },
+    {
+      "epoch": 0.4812252452665723,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 9437
+    },
+    {
+      "epoch": 0.48127623872267766,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 9438
+    },
+    {
+      "epoch": 0.48132723217878304,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 9439
+    },
+    {
+      "epoch": 0.48137822563488847,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 9440
+    },
+    {
+      "epoch": 0.48142921909099384,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 9441
+    },
+    {
+      "epoch": 0.4814802125470992,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 9442
+    },
+    {
+      "epoch": 0.4815312060032046,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 9443
+    },
+    {
+      "epoch": 0.48158219945931,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9444
+    },
+    {
+      "epoch": 0.4816331929154154,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 9445
+    },
+    {
+      "epoch": 0.4816841863715208,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 9446
+    },
+    {
+      "epoch": 0.4817351798276262,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 9447
+    },
+    {
+      "epoch": 0.4817861732837316,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 9448
+    },
+    {
+      "epoch": 0.48183716673983695,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 9449
+    },
+    {
+      "epoch": 0.48188816019594233,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 9450
+    },
+    {
+      "epoch": 0.48193915365204776,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 9451
+    },
+    {
+      "epoch": 0.48199014710815313,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 9452
+    },
+    {
+      "epoch": 0.4820411405642585,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 9453
+    },
+    {
+      "epoch": 0.48209213402036394,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 9454
+    },
+    {
+      "epoch": 0.4821431274764693,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 9455
+    },
+    {
+      "epoch": 0.4821941209325747,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 9456
+    },
+    {
+      "epoch": 0.48224511438868006,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 9457
+    },
+    {
+      "epoch": 0.4822961078447855,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 9458
+    },
+    {
+      "epoch": 0.48234710130089087,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 9459
+    },
+    {
+      "epoch": 0.48239809475699624,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4824490882131017,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 9461
+    },
+    {
+      "epoch": 0.48250008166920705,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 9462
+    },
+    {
+      "epoch": 0.4825510751253124,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 9463
+    },
+    {
+      "epoch": 0.4826020685814178,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 9464
+    },
+    {
+      "epoch": 0.48265306203752323,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 9465
+    },
+    {
+      "epoch": 0.4827040554936286,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 9466
+    },
+    {
+      "epoch": 0.482755048949734,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 9467
+    },
+    {
+      "epoch": 0.4828060424058394,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9468
+    },
+    {
+      "epoch": 0.4828570358619448,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 9469
+    },
+    {
+      "epoch": 0.48290802931805016,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48295902277415553,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 9471
+    },
+    {
+      "epoch": 0.48301001623026096,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 9472
+    },
+    {
+      "epoch": 0.48306100968636634,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 9473
+    },
+    {
+      "epoch": 0.4831120031424717,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 9474
+    },
+    {
+      "epoch": 0.48316299659857714,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 9475
+    },
+    {
+      "epoch": 0.4832139900546825,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 9476
+    },
+    {
+      "epoch": 0.4832649835107879,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 9477
+    },
+    {
+      "epoch": 0.48331597696689327,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 9478
+    },
+    {
+      "epoch": 0.4833669704229987,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 9479
+    },
+    {
+      "epoch": 0.4834179638791041,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48346895733520945,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7796,
+      "step": 9481
+    },
+    {
+      "epoch": 0.4835199507913149,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 9482
+    },
+    {
+      "epoch": 0.48357094424742025,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 9483
+    },
+    {
+      "epoch": 0.48362193770352563,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 9484
+    },
+    {
+      "epoch": 0.483672931159631,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 9485
+    },
+    {
+      "epoch": 0.48372392461573643,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 9486
+    },
+    {
+      "epoch": 0.4837749180718418,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 9487
+    },
+    {
+      "epoch": 0.4838259115279472,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 9488
+    },
+    {
+      "epoch": 0.4838769049840526,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 9489
+    },
+    {
+      "epoch": 0.483927898440158,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48397889189626336,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7661,
+      "step": 9491
+    },
+    {
+      "epoch": 0.48402988535236874,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 9492
+    },
+    {
+      "epoch": 0.48408087880847417,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 9493
+    },
+    {
+      "epoch": 0.48413187226457954,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 9494
+    },
+    {
+      "epoch": 0.4841828657206849,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 9495
+    },
+    {
+      "epoch": 0.48423385917679035,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 9496
+    },
+    {
+      "epoch": 0.4842848526328957,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 9497
+    },
+    {
+      "epoch": 0.4843358460890011,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 9498
+    },
+    {
+      "epoch": 0.4843868395451065,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 9499
+    },
+    {
+      "epoch": 0.4844378330012119,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4844888264573173,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 9501
+    },
+    {
+      "epoch": 0.48453981991342265,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 9502
+    },
+    {
+      "epoch": 0.4845908133695281,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 9503
+    },
+    {
+      "epoch": 0.48464180682563346,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 9504
+    },
+    {
+      "epoch": 0.48469280028173883,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 9505
+    },
+    {
+      "epoch": 0.4847437937378442,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 9506
+    },
+    {
+      "epoch": 0.48479478719394964,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 9507
+    },
+    {
+      "epoch": 0.484845780650055,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7859,
+      "step": 9508
+    },
+    {
+      "epoch": 0.4848967741061604,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 9509
+    },
+    {
+      "epoch": 0.4849477675622658,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4849987610183712,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 9511
+    },
+    {
+      "epoch": 0.48504975447447657,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 9512
+    },
+    {
+      "epoch": 0.48510074793058194,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 9513
+    },
+    {
+      "epoch": 0.4851517413866874,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 9514
+    },
+    {
+      "epoch": 0.48520273484279275,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 9515
+    },
+    {
+      "epoch": 0.4852537282988981,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 9516
+    },
+    {
+      "epoch": 0.48530472175500355,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 9517
+    },
+    {
+      "epoch": 0.48535571521110893,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 9518
+    },
+    {
+      "epoch": 0.4854067086672143,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 9519
+    },
+    {
+      "epoch": 0.4854577021233197,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4855086955794251,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 9521
+    },
+    {
+      "epoch": 0.4855596890355305,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 9522
+    },
+    {
+      "epoch": 0.48561068249163586,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 9523
+    },
+    {
+      "epoch": 0.4856616759477413,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 9524
+    },
+    {
+      "epoch": 0.48571266940384666,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 9525
+    },
+    {
+      "epoch": 0.48576366285995204,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 9526
+    },
+    {
+      "epoch": 0.4858146563160574,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 9527
+    },
+    {
+      "epoch": 0.48586564977216284,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 9528
+    },
+    {
+      "epoch": 0.4859166432282682,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 9529
+    },
+    {
+      "epoch": 0.4859676366843736,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 9530
+    },
+    {
+      "epoch": 0.486018630140479,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 9531
+    },
+    {
+      "epoch": 0.4860696235965844,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 9532
+    },
+    {
+      "epoch": 0.4861206170526898,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 9533
+    },
+    {
+      "epoch": 0.48617161050879515,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 9534
+    },
+    {
+      "epoch": 0.4862226039649006,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 9535
+    },
+    {
+      "epoch": 0.48627359742100595,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 9536
+    },
+    {
+      "epoch": 0.48632459087711133,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 9537
+    },
+    {
+      "epoch": 0.48637558433321676,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 9538
+    },
+    {
+      "epoch": 0.48642657778932213,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 9539
+    },
+    {
+      "epoch": 0.4864775712454275,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4865285647015329,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 9541
+    },
+    {
+      "epoch": 0.4865795581576383,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 9542
+    },
+    {
+      "epoch": 0.4866305516137437,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 9543
+    },
+    {
+      "epoch": 0.48668154506984906,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 9544
+    },
+    {
+      "epoch": 0.4867325385259545,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9545
+    },
+    {
+      "epoch": 0.48678353198205987,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 9546
+    },
+    {
+      "epoch": 0.48683452543816524,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 9547
+    },
+    {
+      "epoch": 0.4868855188942706,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 9548
+    },
+    {
+      "epoch": 0.48693651235037605,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 9549
+    },
+    {
+      "epoch": 0.4869875058064814,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 9550
+    },
+    {
+      "epoch": 0.4870384992625868,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 9551
+    },
+    {
+      "epoch": 0.48708949271869223,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 9552
+    },
+    {
+      "epoch": 0.4871404861747976,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 9553
+    },
+    {
+      "epoch": 0.487191479630903,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 9554
+    },
+    {
+      "epoch": 0.48724247308700835,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4872934665431138,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 9556
+    },
+    {
+      "epoch": 0.48734445999921916,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9557
+    },
+    {
+      "epoch": 0.48739545345532453,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 9558
+    },
+    {
+      "epoch": 0.48744644691142996,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 9559
+    },
+    {
+      "epoch": 0.48749744036753534,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4875484338236407,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 9561
+    },
+    {
+      "epoch": 0.4875994272797461,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 9562
+    },
+    {
+      "epoch": 0.4876504207358515,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 9563
+    },
+    {
+      "epoch": 0.4877014141919569,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 9564
+    },
+    {
+      "epoch": 0.48775240764806227,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7727,
+      "step": 9565
+    },
+    {
+      "epoch": 0.4878034011041677,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 9566
+    },
+    {
+      "epoch": 0.4878543945602731,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 9567
+    },
+    {
+      "epoch": 0.48790538801637845,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 9568
+    },
+    {
+      "epoch": 0.4879563814724838,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 9569
+    },
+    {
+      "epoch": 0.48800737492858925,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8334,
+      "step": 9570
+    },
+    {
+      "epoch": 0.48805836838469463,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 9571
+    },
+    {
+      "epoch": 0.4881093618408,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 9572
+    },
+    {
+      "epoch": 0.48816035529690543,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 9573
+    },
+    {
+      "epoch": 0.4882113487530108,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 9574
+    },
+    {
+      "epoch": 0.4882623422091162,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 9575
+    },
+    {
+      "epoch": 0.48831333566522156,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 9576
+    },
+    {
+      "epoch": 0.488364329121327,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 9577
+    },
+    {
+      "epoch": 0.48841532257743236,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 9578
+    },
+    {
+      "epoch": 0.48846631603353774,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 9579
+    },
+    {
+      "epoch": 0.48851730948964317,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 9580
+    },
+    {
+      "epoch": 0.48856830294574854,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7845,
+      "step": 9581
+    },
+    {
+      "epoch": 0.4886192964018539,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 9582
+    },
+    {
+      "epoch": 0.4886702898579593,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 9583
+    },
+    {
+      "epoch": 0.4887212833140647,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 9584
+    },
+    {
+      "epoch": 0.4887722767701701,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 9585
+    },
+    {
+      "epoch": 0.4888232702262755,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 9586
+    },
+    {
+      "epoch": 0.4888742636823809,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 9587
+    },
+    {
+      "epoch": 0.4889252571384863,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8458,
+      "step": 9588
+    },
+    {
+      "epoch": 0.48897625059459165,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 9589
+    },
+    {
+      "epoch": 0.48902724405069703,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 9590
+    },
+    {
+      "epoch": 0.48907823750680246,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 9591
+    },
+    {
+      "epoch": 0.48912923096290783,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 9592
+    },
+    {
+      "epoch": 0.4891802244190132,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 9593
+    },
+    {
+      "epoch": 0.48923121787511864,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 9594
+    },
+    {
+      "epoch": 0.489282211331224,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8485,
+      "step": 9595
+    },
+    {
+      "epoch": 0.4893332047873294,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 9596
+    },
+    {
+      "epoch": 0.48938419824343476,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 9597
+    },
+    {
+      "epoch": 0.4894351916995402,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 9598
+    },
+    {
+      "epoch": 0.48948618515564557,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 9599
+    },
+    {
+      "epoch": 0.48953717861175094,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4895881720678564,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 9601
+    },
+    {
+      "epoch": 0.48963916552396175,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 9602
+    },
+    {
+      "epoch": 0.4896901589800671,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 9603
+    },
+    {
+      "epoch": 0.4897411524361725,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 9604
+    },
+    {
+      "epoch": 0.48979214589227793,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 9605
+    },
+    {
+      "epoch": 0.4898431393483833,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 9606
+    },
+    {
+      "epoch": 0.4898941328044887,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 9607
+    },
+    {
+      "epoch": 0.4899451262605941,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 9608
+    },
+    {
+      "epoch": 0.4899961197166995,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 9609
+    },
+    {
+      "epoch": 0.49004711317280486,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 9610
+    },
+    {
+      "epoch": 0.49009810662891024,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 9611
+    },
+    {
+      "epoch": 0.49014910008501567,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 9612
+    },
+    {
+      "epoch": 0.49020009354112104,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 9613
+    },
+    {
+      "epoch": 0.4902510869972264,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 9614
+    },
+    {
+      "epoch": 0.49030208045333185,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 9615
+    },
+    {
+      "epoch": 0.4903530739094372,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 9616
+    },
+    {
+      "epoch": 0.4904040673655426,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 9617
+    },
+    {
+      "epoch": 0.49045506082164797,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 9618
+    },
+    {
+      "epoch": 0.4905060542777534,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 9619
+    },
+    {
+      "epoch": 0.4905570477338588,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49060804118996415,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 9621
+    },
+    {
+      "epoch": 0.4906590346460696,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 9622
+    },
+    {
+      "epoch": 0.49071002810217496,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 9623
+    },
+    {
+      "epoch": 0.49076102155828033,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 9624
+    },
+    {
+      "epoch": 0.4908120150143857,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 9625
+    },
+    {
+      "epoch": 0.49086300847049114,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 9626
+    },
+    {
+      "epoch": 0.4909140019265965,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 9627
+    },
+    {
+      "epoch": 0.4909649953827019,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9628
+    },
+    {
+      "epoch": 0.4910159888388073,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 9629
+    },
+    {
+      "epoch": 0.4910669822949127,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 9630
+    },
+    {
+      "epoch": 0.49111797575101807,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 9631
+    },
+    {
+      "epoch": 0.49116896920712344,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 9632
+    },
+    {
+      "epoch": 0.49121996266322887,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 9633
+    },
+    {
+      "epoch": 0.49127095611933425,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 9634
+    },
+    {
+      "epoch": 0.4913219495754396,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 9635
+    },
+    {
+      "epoch": 0.49137294303154505,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8555,
+      "step": 9636
+    },
+    {
+      "epoch": 0.4914239364876504,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 9637
+    },
+    {
+      "epoch": 0.4914749299437558,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 9638
+    },
+    {
+      "epoch": 0.4915259233998612,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 9639
+    },
+    {
+      "epoch": 0.4915769168559666,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 9640
+    },
+    {
+      "epoch": 0.491627910312072,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 9641
+    },
+    {
+      "epoch": 0.49167890376817736,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 9642
+    },
+    {
+      "epoch": 0.4917298972242828,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 9643
+    },
+    {
+      "epoch": 0.49178089068038816,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 9644
+    },
+    {
+      "epoch": 0.49183188413649354,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 9645
+    },
+    {
+      "epoch": 0.4918828775925989,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7754,
+      "step": 9646
+    },
+    {
+      "epoch": 0.49193387104870434,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 9647
+    },
+    {
+      "epoch": 0.4919848645048097,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 9648
+    },
+    {
+      "epoch": 0.4920358579609151,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 9649
+    },
+    {
+      "epoch": 0.4920868514170205,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4921378448731259,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 9651
+    },
+    {
+      "epoch": 0.49218883832923127,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 9652
+    },
+    {
+      "epoch": 0.49223983178533665,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 9653
+    },
+    {
+      "epoch": 0.4922908252414421,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 9654
+    },
+    {
+      "epoch": 0.49234181869754745,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 9655
+    },
+    {
+      "epoch": 0.4923928121536528,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 9656
+    },
+    {
+      "epoch": 0.49244380560975826,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8493,
+      "step": 9657
+    },
+    {
+      "epoch": 0.49249479906586363,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 9658
+    },
+    {
+      "epoch": 0.492545792521969,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 9659
+    },
+    {
+      "epoch": 0.4925967859780744,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4926477794341798,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 9661
+    },
+    {
+      "epoch": 0.4926987728902852,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 9662
+    },
+    {
+      "epoch": 0.49274976634639056,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 9663
+    },
+    {
+      "epoch": 0.492800759802496,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 9664
+    },
+    {
+      "epoch": 0.49285175325860137,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 9665
+    },
+    {
+      "epoch": 0.49290274671470674,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 9666
+    },
+    {
+      "epoch": 0.4929537401708121,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 9667
+    },
+    {
+      "epoch": 0.49300473362691755,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9668
+    },
+    {
+      "epoch": 0.4930557270830229,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9669
+    },
+    {
+      "epoch": 0.4931067205391283,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4931577139952337,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 9671
+    },
+    {
+      "epoch": 0.4932087074513391,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8496,
+      "step": 9672
+    },
+    {
+      "epoch": 0.4932597009074445,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 9673
+    },
+    {
+      "epoch": 0.49331069436354985,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 9674
+    },
+    {
+      "epoch": 0.4933616878196553,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 9675
+    },
+    {
+      "epoch": 0.49341268127576066,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9676
+    },
+    {
+      "epoch": 0.49346367473186603,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7633,
+      "step": 9677
+    },
+    {
+      "epoch": 0.49351466818797146,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 9678
+    },
+    {
+      "epoch": 0.49356566164407684,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 9679
+    },
+    {
+      "epoch": 0.4936166551001822,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7689,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4936676485562876,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 9681
+    },
+    {
+      "epoch": 0.493718642012393,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 9682
+    },
+    {
+      "epoch": 0.4937696354684984,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 9683
+    },
+    {
+      "epoch": 0.49382062892460377,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 9684
+    },
+    {
+      "epoch": 0.4938716223807092,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 9685
+    },
+    {
+      "epoch": 0.49392261583681457,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 9686
+    },
+    {
+      "epoch": 0.49397360929291995,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 9687
+    },
+    {
+      "epoch": 0.4940246027490253,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 9688
+    },
+    {
+      "epoch": 0.49407559620513075,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 9689
+    },
+    {
+      "epoch": 0.4941265896612361,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4941775831173415,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 9691
+    },
+    {
+      "epoch": 0.49422857657344693,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 9692
+    },
+    {
+      "epoch": 0.4942795700295523,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 9693
+    },
+    {
+      "epoch": 0.4943305634856577,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 9694
+    },
+    {
+      "epoch": 0.49438155694176306,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 9695
+    },
+    {
+      "epoch": 0.4944325503978685,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 9696
+    },
+    {
+      "epoch": 0.49448354385397386,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 9697
+    },
+    {
+      "epoch": 0.49453453731007924,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9698
+    },
+    {
+      "epoch": 0.49458553076618467,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 9699
+    },
+    {
+      "epoch": 0.49463652422229004,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4946875176783954,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 9701
+    },
+    {
+      "epoch": 0.4947385111345008,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 9702
+    },
+    {
+      "epoch": 0.4947895045906062,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 9703
+    },
+    {
+      "epoch": 0.4948404980467116,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 9704
+    },
+    {
+      "epoch": 0.49489149150281697,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7762,
+      "step": 9705
+    },
+    {
+      "epoch": 0.4949424849589224,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 9706
+    },
+    {
+      "epoch": 0.4949934784150278,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 9707
+    },
+    {
+      "epoch": 0.49504447187113315,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 9708
+    },
+    {
+      "epoch": 0.4950954653272385,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 9709
+    },
+    {
+      "epoch": 0.49514645878334396,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 9710
+    },
+    {
+      "epoch": 0.49519745223944933,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 9711
+    },
+    {
+      "epoch": 0.4952484456955547,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 9712
+    },
+    {
+      "epoch": 0.49529943915166014,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 9713
+    },
+    {
+      "epoch": 0.4953504326077655,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 9714
+    },
+    {
+      "epoch": 0.4954014260638709,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8491,
+      "step": 9715
+    },
+    {
+      "epoch": 0.49545241951997626,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 9716
+    },
+    {
+      "epoch": 0.4955034129760817,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 9717
+    },
+    {
+      "epoch": 0.49555440643218707,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 9718
+    },
+    {
+      "epoch": 0.49560539988829244,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 9719
+    },
+    {
+      "epoch": 0.49565639334439787,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 9720
+    },
+    {
+      "epoch": 0.49570738680050325,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 9721
+    },
+    {
+      "epoch": 0.4957583802566086,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 9722
+    },
+    {
+      "epoch": 0.495809373712714,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 9723
+    },
+    {
+      "epoch": 0.4958603671688194,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 9724
+    },
+    {
+      "epoch": 0.4959113606249248,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.841,
+      "step": 9725
+    },
+    {
+      "epoch": 0.4959623540810302,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 9726
+    },
+    {
+      "epoch": 0.4960133475371356,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 9727
+    },
+    {
+      "epoch": 0.496064340993241,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 9728
+    },
+    {
+      "epoch": 0.49611533444934636,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 9729
+    },
+    {
+      "epoch": 0.49616632790545173,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49621732136155716,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 9731
+    },
+    {
+      "epoch": 0.49626831481766254,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7818,
+      "step": 9732
+    },
+    {
+      "epoch": 0.4963193082737679,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7773,
+      "step": 9733
+    },
+    {
+      "epoch": 0.49637030172987334,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 9734
+    },
+    {
+      "epoch": 0.4964212951859787,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 9735
+    },
+    {
+      "epoch": 0.4964722886420841,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 9736
+    },
+    {
+      "epoch": 0.49652328209818947,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 9737
+    },
+    {
+      "epoch": 0.4965742755542949,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8548,
+      "step": 9738
+    },
+    {
+      "epoch": 0.49662526901040027,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 9739
+    },
+    {
+      "epoch": 0.49667626246650565,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4967272559226111,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 9741
+    },
+    {
+      "epoch": 0.49677824937871645,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 9742
+    },
+    {
+      "epoch": 0.4968292428348218,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9743
+    },
+    {
+      "epoch": 0.4968802362909272,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 9744
+    },
+    {
+      "epoch": 0.49693122974703263,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 9745
+    },
+    {
+      "epoch": 0.496982223203138,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 9746
+    },
+    {
+      "epoch": 0.4970332166592434,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 9747
+    },
+    {
+      "epoch": 0.4970842101153488,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 9748
+    },
+    {
+      "epoch": 0.4971352035714542,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 9749
+    },
+    {
+      "epoch": 0.49718619702755956,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 9750
+    },
+    {
+      "epoch": 0.49723719048366494,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 9751
+    },
+    {
+      "epoch": 0.49728818393977037,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 9752
+    },
+    {
+      "epoch": 0.49733917739587574,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 9753
+    },
+    {
+      "epoch": 0.4973901708519811,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 9754
+    },
+    {
+      "epoch": 0.49744116430808655,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 9755
+    },
+    {
+      "epoch": 0.4974921577641919,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 9756
+    },
+    {
+      "epoch": 0.4975431512202973,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 9757
+    },
+    {
+      "epoch": 0.49759414467640267,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 9758
+    },
+    {
+      "epoch": 0.4976451381325081,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 9759
+    },
+    {
+      "epoch": 0.4976961315886135,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 9760
+    },
+    {
+      "epoch": 0.49774712504471885,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 9761
+    },
+    {
+      "epoch": 0.4977981185008243,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 9762
+    },
+    {
+      "epoch": 0.49784911195692966,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 9763
+    },
+    {
+      "epoch": 0.49790010541303503,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 9764
+    },
+    {
+      "epoch": 0.4979510988691404,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 9765
+    },
+    {
+      "epoch": 0.49800209232524584,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 9766
+    },
+    {
+      "epoch": 0.4980530857813512,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 9767
+    },
+    {
+      "epoch": 0.4981040792374566,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 9768
+    },
+    {
+      "epoch": 0.498155072693562,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 9769
+    },
+    {
+      "epoch": 0.4982060661496674,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 9770
+    },
+    {
+      "epoch": 0.49825705960577277,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 9771
+    },
+    {
+      "epoch": 0.49830805306187814,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 9772
+    },
+    {
+      "epoch": 0.4983590465179836,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 9773
+    },
+    {
+      "epoch": 0.49841003997408895,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 9774
+    },
+    {
+      "epoch": 0.4984610334301943,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 9775
+    },
+    {
+      "epoch": 0.49851202688629975,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 9776
+    },
+    {
+      "epoch": 0.4985630203424051,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 9777
+    },
+    {
+      "epoch": 0.4986140137985105,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 9778
+    },
+    {
+      "epoch": 0.4986650072546159,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 9779
+    },
+    {
+      "epoch": 0.4987160007107213,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4987669941668267,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 9781
+    },
+    {
+      "epoch": 0.49881798762293206,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 9782
+    },
+    {
+      "epoch": 0.4988689810790375,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 9783
+    },
+    {
+      "epoch": 0.49891997453514286,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7579,
+      "step": 9784
+    },
+    {
+      "epoch": 0.49897096799124824,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 9785
+    },
+    {
+      "epoch": 0.4990219614473536,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 9786
+    },
+    {
+      "epoch": 0.49907295490345904,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 9787
+    },
+    {
+      "epoch": 0.4991239483595644,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 9788
+    },
+    {
+      "epoch": 0.4991749418156698,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 9789
+    },
+    {
+      "epoch": 0.4992259352717752,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4992769287278806,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.7591,
+      "step": 9791
+    },
+    {
+      "epoch": 0.499327922183986,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 9792
+    },
+    {
+      "epoch": 0.49937891564009135,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 9793
+    },
+    {
+      "epoch": 0.4994299090961968,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 9794
+    },
+    {
+      "epoch": 0.49948090255230215,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 9795
+    },
+    {
+      "epoch": 0.4995318960084075,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 9796
+    },
+    {
+      "epoch": 0.49958288946451296,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 9797
+    },
+    {
+      "epoch": 0.49963388292061833,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 9798
+    },
+    {
+      "epoch": 0.4996848763767237,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 9799
+    },
+    {
+      "epoch": 0.4997358698328291,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 9800
+    },
+    {
+      "epoch": 0.4997868632889345,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 9801
+    },
+    {
+      "epoch": 0.4998378567450399,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 9802
+    },
+    {
+      "epoch": 0.49988885020114526,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 9803
+    },
+    {
+      "epoch": 0.4999398436572507,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 9804
+    },
+    {
+      "epoch": 0.49999083711335607,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 9805
+    },
+    {
+      "epoch": 0.5000418305694615,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 9806
+    },
+    {
+      "epoch": 0.5000928240255669,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9807
+    },
+    {
+      "epoch": 0.5001438174816722,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 9808
+    },
+    {
+      "epoch": 0.5001948109377776,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 9809
+    },
+    {
+      "epoch": 0.500245804393883,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5002967978499884,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 9811
+    },
+    {
+      "epoch": 0.5003477913060937,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 9812
+    },
+    {
+      "epoch": 0.5003987847621992,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 9813
+    },
+    {
+      "epoch": 0.5004497782183046,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 9814
+    },
+    {
+      "epoch": 0.50050077167441,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 9815
+    },
+    {
+      "epoch": 0.5005517651305154,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 9816
+    },
+    {
+      "epoch": 0.5006027585866207,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 9817
+    },
+    {
+      "epoch": 0.5006537520427261,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 9818
+    },
+    {
+      "epoch": 0.5007047454988315,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 9819
+    },
+    {
+      "epoch": 0.500755738954937,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5008067324110423,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 9821
+    },
+    {
+      "epoch": 0.5008577258671477,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 9822
+    },
+    {
+      "epoch": 0.5009087193232531,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 9823
+    },
+    {
+      "epoch": 0.5009597127793585,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 9824
+    },
+    {
+      "epoch": 0.5010107062354638,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.855,
+      "step": 9825
+    },
+    {
+      "epoch": 0.5010616996915692,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 9826
+    },
+    {
+      "epoch": 0.5011126931476747,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 9827
+    },
+    {
+      "epoch": 0.5011636866037801,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 9828
+    },
+    {
+      "epoch": 0.5012146800598855,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 9829
+    },
+    {
+      "epoch": 0.5012656735159908,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5013166669720962,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 9831
+    },
+    {
+      "epoch": 0.5013676604282016,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 9832
+    },
+    {
+      "epoch": 0.501418653884307,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 9833
+    },
+    {
+      "epoch": 0.5014696473404124,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 9834
+    },
+    {
+      "epoch": 0.5015206407965178,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 9835
+    },
+    {
+      "epoch": 0.5015716342526232,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 9836
+    },
+    {
+      "epoch": 0.5016226277087286,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 9837
+    },
+    {
+      "epoch": 0.5016736211648339,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 9838
+    },
+    {
+      "epoch": 0.5017246146209393,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 9839
+    },
+    {
+      "epoch": 0.5017756080770447,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5018266015331502,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 9841
+    },
+    {
+      "epoch": 0.5018775949892555,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 9842
+    },
+    {
+      "epoch": 0.5019285884453609,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 9843
+    },
+    {
+      "epoch": 0.5019795819014663,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 9844
+    },
+    {
+      "epoch": 0.5020305753575717,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 9845
+    },
+    {
+      "epoch": 0.502081568813677,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 9846
+    },
+    {
+      "epoch": 0.5021325622697824,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8582,
+      "step": 9847
+    },
+    {
+      "epoch": 0.5021835557258879,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 9848
+    },
+    {
+      "epoch": 0.5022345491819933,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 9849
+    },
+    {
+      "epoch": 0.5022855426380987,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 9850
+    },
+    {
+      "epoch": 0.502336536094204,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 9851
+    },
+    {
+      "epoch": 0.5023875295503094,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 9852
+    },
+    {
+      "epoch": 0.5024385230064148,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 9853
+    },
+    {
+      "epoch": 0.5024895164625202,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 9854
+    },
+    {
+      "epoch": 0.5025405099186256,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 9855
+    },
+    {
+      "epoch": 0.502591503374731,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 9856
+    },
+    {
+      "epoch": 0.5026424968308364,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 9857
+    },
+    {
+      "epoch": 0.5026934902869418,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 9858
+    },
+    {
+      "epoch": 0.5027444837430471,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 9859
+    },
+    {
+      "epoch": 0.5027954771991525,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5028464706552579,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 9861
+    },
+    {
+      "epoch": 0.5028974641113634,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 9862
+    },
+    {
+      "epoch": 0.5029484575674688,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 9863
+    },
+    {
+      "epoch": 0.5029994510235741,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8426,
+      "step": 9864
+    },
+    {
+      "epoch": 0.5030504444796795,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 9865
+    },
+    {
+      "epoch": 0.5031014379357849,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 9866
+    },
+    {
+      "epoch": 0.5031524313918903,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 9867
+    },
+    {
+      "epoch": 0.5032034248479956,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 9868
+    },
+    {
+      "epoch": 0.5032544183041011,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8447,
+      "step": 9869
+    },
+    {
+      "epoch": 0.5033054117602065,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5033564052163119,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 9871
+    },
+    {
+      "epoch": 0.5034073986724172,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 9872
+    },
+    {
+      "epoch": 0.5034583921285226,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 9873
+    },
+    {
+      "epoch": 0.503509385584628,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 9874
+    },
+    {
+      "epoch": 0.5035603790407334,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 9875
+    },
+    {
+      "epoch": 0.5036113724968388,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 9876
+    },
+    {
+      "epoch": 0.5036623659529442,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 9877
+    },
+    {
+      "epoch": 0.5037133594090496,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 9878
+    },
+    {
+      "epoch": 0.503764352865155,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 9879
+    },
+    {
+      "epoch": 0.5038153463212603,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5038663397773657,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 9881
+    },
+    {
+      "epoch": 0.5039173332334711,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 9882
+    },
+    {
+      "epoch": 0.5039683266895766,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.837,
+      "step": 9883
+    },
+    {
+      "epoch": 0.504019320145682,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 9884
+    },
+    {
+      "epoch": 0.5040703136017873,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 9885
+    },
+    {
+      "epoch": 0.5041213070578927,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 9886
+    },
+    {
+      "epoch": 0.5041723005139981,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 9887
+    },
+    {
+      "epoch": 0.5042232939701035,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 9888
+    },
+    {
+      "epoch": 0.5042742874262088,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 9889
+    },
+    {
+      "epoch": 0.5043252808823143,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5043762743384197,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 9891
+    },
+    {
+      "epoch": 0.5044272677945251,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9892
+    },
+    {
+      "epoch": 0.5044782612506304,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 9893
+    },
+    {
+      "epoch": 0.5045292547067358,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 9894
+    },
+    {
+      "epoch": 0.5045802481628412,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 9895
+    },
+    {
+      "epoch": 0.5046312416189466,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9896
+    },
+    {
+      "epoch": 0.504682235075052,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 9897
+    },
+    {
+      "epoch": 0.5047332285311574,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 9898
+    },
+    {
+      "epoch": 0.5047842219872628,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 9899
+    },
+    {
+      "epoch": 0.5048352154433682,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5048862088994736,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 9901
+    },
+    {
+      "epoch": 0.5049372023555789,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 9902
+    },
+    {
+      "epoch": 0.5049881958116843,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 9903
+    },
+    {
+      "epoch": 0.5050391892677898,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 9904
+    },
+    {
+      "epoch": 0.5050901827238952,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 9905
+    },
+    {
+      "epoch": 0.5051411761800005,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8454,
+      "step": 9906
+    },
+    {
+      "epoch": 0.5051921696361059,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 9907
+    },
+    {
+      "epoch": 0.5052431630922113,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7761,
+      "step": 9908
+    },
+    {
+      "epoch": 0.5052941565483167,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 9909
+    },
+    {
+      "epoch": 0.505345150004422,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5053961434605275,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 9911
+    },
+    {
+      "epoch": 0.5054471369166329,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 9912
+    },
+    {
+      "epoch": 0.5054981303727383,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 9913
+    },
+    {
+      "epoch": 0.5055491238288436,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 9914
+    },
+    {
+      "epoch": 0.505600117284949,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8524,
+      "step": 9915
+    },
+    {
+      "epoch": 0.5056511107410544,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8452,
+      "step": 9916
+    },
+    {
+      "epoch": 0.5057021041971598,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9917
+    },
+    {
+      "epoch": 0.5057530976532653,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 9918
+    },
+    {
+      "epoch": 0.5058040911093706,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 9919
+    },
+    {
+      "epoch": 0.505855084565476,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5059060780215814,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 9921
+    },
+    {
+      "epoch": 0.5059570714776868,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 9922
+    },
+    {
+      "epoch": 0.5060080649337921,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 9923
+    },
+    {
+      "epoch": 0.5060590583898975,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 9924
+    },
+    {
+      "epoch": 0.506110051846003,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 9925
+    },
+    {
+      "epoch": 0.5061610453021084,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 9926
+    },
+    {
+      "epoch": 0.5062120387582137,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 9927
+    },
+    {
+      "epoch": 0.5062630322143191,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 9928
+    },
+    {
+      "epoch": 0.5063140256704245,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 9929
+    },
+    {
+      "epoch": 0.5063650191265299,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5064160125826352,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 9931
+    },
+    {
+      "epoch": 0.5064670060387407,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 9932
+    },
+    {
+      "epoch": 0.5065179994948461,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7769,
+      "step": 9933
+    },
+    {
+      "epoch": 0.5065689929509515,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8355,
+      "step": 9934
+    },
+    {
+      "epoch": 0.5066199864070569,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7771,
+      "step": 9935
+    },
+    {
+      "epoch": 0.5066709798631622,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7815,
+      "step": 9936
+    },
+    {
+      "epoch": 0.5067219733192676,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 9937
+    },
+    {
+      "epoch": 0.506772966775373,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 9938
+    },
+    {
+      "epoch": 0.5068239602314785,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 9939
+    },
+    {
+      "epoch": 0.5068749536875838,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5069259471436892,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9941
+    },
+    {
+      "epoch": 0.5069769405997946,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 9942
+    },
+    {
+      "epoch": 0.5070279340559,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 9943
+    },
+    {
+      "epoch": 0.5070789275120053,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 9944
+    },
+    {
+      "epoch": 0.5071299209681107,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 9945
+    },
+    {
+      "epoch": 0.5071809144242162,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8459,
+      "step": 9946
+    },
+    {
+      "epoch": 0.5072319078803216,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 9947
+    },
+    {
+      "epoch": 0.507282901336427,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8743,
+      "step": 9948
+    },
+    {
+      "epoch": 0.5073338947925323,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 9949
+    },
+    {
+      "epoch": 0.5073848882486377,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 9950
+    },
+    {
+      "epoch": 0.5074358817047431,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 9951
+    },
+    {
+      "epoch": 0.5074868751608484,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 9952
+    },
+    {
+      "epoch": 0.5075378686169539,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 9953
+    },
+    {
+      "epoch": 0.5075888620730593,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 9954
+    },
+    {
+      "epoch": 0.5076398555291647,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 9955
+    },
+    {
+      "epoch": 0.5076908489852701,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 9956
+    },
+    {
+      "epoch": 0.5077418424413754,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 9957
+    },
+    {
+      "epoch": 0.5077928358974808,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 9958
+    },
+    {
+      "epoch": 0.5078438293535862,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 9959
+    },
+    {
+      "epoch": 0.5078948228096917,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7753,
+      "step": 9960
+    },
+    {
+      "epoch": 0.507945816265797,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 9961
+    },
+    {
+      "epoch": 0.5079968097219024,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 9962
+    },
+    {
+      "epoch": 0.5080478031780078,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 9963
+    },
+    {
+      "epoch": 0.5080987966341132,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 9964
+    },
+    {
+      "epoch": 0.5081497900902185,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 9965
+    },
+    {
+      "epoch": 0.5082007835463239,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 9966
+    },
+    {
+      "epoch": 0.5082517770024294,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 9967
+    },
+    {
+      "epoch": 0.5083027704585348,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.854,
+      "step": 9968
+    },
+    {
+      "epoch": 0.5083537639146402,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7744,
+      "step": 9969
+    },
+    {
+      "epoch": 0.5084047573707455,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5084557508268509,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 9971
+    },
+    {
+      "epoch": 0.5085067442829563,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 9972
+    },
+    {
+      "epoch": 0.5085577377390617,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 9973
+    },
+    {
+      "epoch": 0.5086087311951671,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 9974
+    },
+    {
+      "epoch": 0.5086597246512725,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 9975
+    },
+    {
+      "epoch": 0.5087107181073779,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 9976
+    },
+    {
+      "epoch": 0.5087617115634833,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 9977
+    },
+    {
+      "epoch": 0.5088127050195886,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 9978
+    },
+    {
+      "epoch": 0.508863698475694,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 9979
+    },
+    {
+      "epoch": 0.5089146919317994,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5089656853879049,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 9981
+    },
+    {
+      "epoch": 0.5090166788440103,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 9982
+    },
+    {
+      "epoch": 0.5090676723001156,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 9983
+    },
+    {
+      "epoch": 0.509118665756221,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 9984
+    },
+    {
+      "epoch": 0.5091696592123264,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 9985
+    },
+    {
+      "epoch": 0.5092206526684318,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 9986
+    },
+    {
+      "epoch": 0.5092716461245371,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 9987
+    },
+    {
+      "epoch": 0.5093226395806426,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 9988
+    },
+    {
+      "epoch": 0.509373633036748,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 9989
+    },
+    {
+      "epoch": 0.5094246264928534,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.77,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5094756199489587,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 9991
+    },
+    {
+      "epoch": 0.5095266134050641,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 9992
+    },
+    {
+      "epoch": 0.5095776068611695,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 9993
+    },
+    {
+      "epoch": 0.5096286003172749,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 9994
+    },
+    {
+      "epoch": 0.5096795937733803,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 9995
+    },
+    {
+      "epoch": 0.5097305872294857,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 9996
+    },
+    {
+      "epoch": 0.5097815806855911,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 9997
+    },
+    {
+      "epoch": 0.5098325741416965,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 9998
+    },
+    {
+      "epoch": 0.5098835675978018,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 9999
+    },
+    {
+      "epoch": 0.5099345610539072,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5099855545100126,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 10001
+    },
+    {
+      "epoch": 0.5100365479661181,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 10002
+    },
+    {
+      "epoch": 0.5100875414222235,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 10003
+    },
+    {
+      "epoch": 0.5101385348783288,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 10004
+    },
+    {
+      "epoch": 0.5101895283344342,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 10005
+    },
+    {
+      "epoch": 0.5102405217905396,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 10006
+    },
+    {
+      "epoch": 0.510291515246645,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 10007
+    },
+    {
+      "epoch": 0.5103425087027503,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 10008
+    },
+    {
+      "epoch": 0.5103935021588558,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 10009
+    },
+    {
+      "epoch": 0.5104444956149612,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 10010
+    },
+    {
+      "epoch": 0.5104954890710666,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 10011
+    },
+    {
+      "epoch": 0.5105464825271719,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 10012
+    },
+    {
+      "epoch": 0.5105974759832773,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 10013
+    },
+    {
+      "epoch": 0.5106484694393827,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 10014
+    },
+    {
+      "epoch": 0.5106994628954881,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 10015
+    },
+    {
+      "epoch": 0.5107504563515936,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 10016
+    },
+    {
+      "epoch": 0.5108014498076989,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 10017
+    },
+    {
+      "epoch": 0.5108524432638043,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8442,
+      "step": 10018
+    },
+    {
+      "epoch": 0.5109034367199097,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10019
+    },
+    {
+      "epoch": 0.510954430176015,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 10020
+    },
+    {
+      "epoch": 0.5110054236321204,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 10021
+    },
+    {
+      "epoch": 0.5110564170882258,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 10022
+    },
+    {
+      "epoch": 0.5111074105443313,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 10023
+    },
+    {
+      "epoch": 0.5111584040004367,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 10024
+    },
+    {
+      "epoch": 0.511209397456542,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10025
+    },
+    {
+      "epoch": 0.5112603909126474,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 10026
+    },
+    {
+      "epoch": 0.5113113843687528,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 10027
+    },
+    {
+      "epoch": 0.5113623778248582,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 10028
+    },
+    {
+      "epoch": 0.5114133712809635,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 10029
+    },
+    {
+      "epoch": 0.511464364737069,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10030
+    },
+    {
+      "epoch": 0.5115153581931744,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 10031
+    },
+    {
+      "epoch": 0.5115663516492798,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 10032
+    },
+    {
+      "epoch": 0.5116173451053851,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 10033
+    },
+    {
+      "epoch": 0.5116683385614905,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 10034
+    },
+    {
+      "epoch": 0.5117193320175959,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 10035
+    },
+    {
+      "epoch": 0.5117703254737013,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10036
+    },
+    {
+      "epoch": 0.5118213189298068,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 10037
+    },
+    {
+      "epoch": 0.5118723123859121,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8423,
+      "step": 10038
+    },
+    {
+      "epoch": 0.5119233058420175,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 10039
+    },
+    {
+      "epoch": 0.5119742992981229,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 10040
+    },
+    {
+      "epoch": 0.5120252927542283,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 10041
+    },
+    {
+      "epoch": 0.5120762862103336,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 10042
+    },
+    {
+      "epoch": 0.512127279666439,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 10043
+    },
+    {
+      "epoch": 0.5121782731225445,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 10044
+    },
+    {
+      "epoch": 0.5122292665786499,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 10045
+    },
+    {
+      "epoch": 0.5122802600347552,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 10046
+    },
+    {
+      "epoch": 0.5123312534908606,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10047
+    },
+    {
+      "epoch": 0.512382246946966,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 10048
+    },
+    {
+      "epoch": 0.5124332404030714,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 10049
+    },
+    {
+      "epoch": 0.5124842338591767,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8369,
+      "step": 10050
+    },
+    {
+      "epoch": 0.5125352273152822,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 10051
+    },
+    {
+      "epoch": 0.5125862207713876,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 10052
+    },
+    {
+      "epoch": 0.512637214227493,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 10053
+    },
+    {
+      "epoch": 0.5126882076835984,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 10054
+    },
+    {
+      "epoch": 0.5127392011397037,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 10055
+    },
+    {
+      "epoch": 0.5127901945958091,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 10056
+    },
+    {
+      "epoch": 0.5128411880519145,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10057
+    },
+    {
+      "epoch": 0.51289218150802,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 10058
+    },
+    {
+      "epoch": 0.5129431749641253,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 10059
+    },
+    {
+      "epoch": 0.5129941684202307,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5130451618763361,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 10061
+    },
+    {
+      "epoch": 0.5130961553324415,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 10062
+    },
+    {
+      "epoch": 0.5131471487885468,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 10063
+    },
+    {
+      "epoch": 0.5131981422446522,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 10064
+    },
+    {
+      "epoch": 0.5132491357007577,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 10065
+    },
+    {
+      "epoch": 0.5133001291568631,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 10066
+    },
+    {
+      "epoch": 0.5133511226129684,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 10067
+    },
+    {
+      "epoch": 0.5134021160690738,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10068
+    },
+    {
+      "epoch": 0.5134531095251792,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 10069
+    },
+    {
+      "epoch": 0.5135041029812846,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 10070
+    },
+    {
+      "epoch": 0.51355509643739,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10071
+    },
+    {
+      "epoch": 0.5136060898934954,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 10072
+    },
+    {
+      "epoch": 0.5136570833496008,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 10073
+    },
+    {
+      "epoch": 0.5137080768057062,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 10074
+    },
+    {
+      "epoch": 0.5137590702618116,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 10075
+    },
+    {
+      "epoch": 0.5138100637179169,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 10076
+    },
+    {
+      "epoch": 0.5138610571740223,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 10077
+    },
+    {
+      "epoch": 0.5139120506301277,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 10078
+    },
+    {
+      "epoch": 0.5139630440862332,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 10079
+    },
+    {
+      "epoch": 0.5140140375423385,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5140650309984439,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10081
+    },
+    {
+      "epoch": 0.5141160244545493,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 10082
+    },
+    {
+      "epoch": 0.5141670179106547,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 10083
+    },
+    {
+      "epoch": 0.51421801136676,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10084
+    },
+    {
+      "epoch": 0.5142690048228654,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 10085
+    },
+    {
+      "epoch": 0.5143199982789709,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 10086
+    },
+    {
+      "epoch": 0.5143709917350763,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 10087
+    },
+    {
+      "epoch": 0.5144219851911817,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 10088
+    },
+    {
+      "epoch": 0.514472978647287,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10089
+    },
+    {
+      "epoch": 0.5145239721033924,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5145749655594978,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 10091
+    },
+    {
+      "epoch": 0.5146259590156032,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 10092
+    },
+    {
+      "epoch": 0.5146769524717086,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 10093
+    },
+    {
+      "epoch": 0.514727945927814,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8363,
+      "step": 10094
+    },
+    {
+      "epoch": 0.5147789393839194,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 10095
+    },
+    {
+      "epoch": 0.5148299328400248,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 10096
+    },
+    {
+      "epoch": 0.5148809262961301,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 10097
+    },
+    {
+      "epoch": 0.5149319197522355,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 10098
+    },
+    {
+      "epoch": 0.5149829132083409,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 10099
+    },
+    {
+      "epoch": 0.5150339066644464,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5150849001205517,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 10101
+    },
+    {
+      "epoch": 0.5151358935766571,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 10102
+    },
+    {
+      "epoch": 0.5151868870327625,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10103
+    },
+    {
+      "epoch": 0.5152378804888679,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 10104
+    },
+    {
+      "epoch": 0.5152888739449732,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 10105
+    },
+    {
+      "epoch": 0.5153398674010786,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 10106
+    },
+    {
+      "epoch": 0.5153908608571841,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 10107
+    },
+    {
+      "epoch": 0.5154418543132895,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 10108
+    },
+    {
+      "epoch": 0.5154928477693949,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 10109
+    },
+    {
+      "epoch": 0.5155438412255002,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5155948346816056,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 10111
+    },
+    {
+      "epoch": 0.515645828137711,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 10112
+    },
+    {
+      "epoch": 0.5156968215938164,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7813,
+      "step": 10113
+    },
+    {
+      "epoch": 0.5157478150499218,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 10114
+    },
+    {
+      "epoch": 0.5157988085060272,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 10115
+    },
+    {
+      "epoch": 0.5158498019621326,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 10116
+    },
+    {
+      "epoch": 0.515900795418238,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 10117
+    },
+    {
+      "epoch": 0.5159517888743433,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 10118
+    },
+    {
+      "epoch": 0.5160027823304487,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 10119
+    },
+    {
+      "epoch": 0.5160537757865541,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 10120
+    },
+    {
+      "epoch": 0.5161047692426596,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 10121
+    },
+    {
+      "epoch": 0.516155762698765,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8479,
+      "step": 10122
+    },
+    {
+      "epoch": 0.5162067561548703,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 10123
+    },
+    {
+      "epoch": 0.5162577496109757,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 10124
+    },
+    {
+      "epoch": 0.5163087430670811,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10125
+    },
+    {
+      "epoch": 0.5163597365231865,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10126
+    },
+    {
+      "epoch": 0.5164107299792918,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 10127
+    },
+    {
+      "epoch": 0.5164617234353973,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 10128
+    },
+    {
+      "epoch": 0.5165127168915027,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7706,
+      "step": 10129
+    },
+    {
+      "epoch": 0.5165637103476081,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5166147038037134,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 10131
+    },
+    {
+      "epoch": 0.5166656972598188,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 10132
+    },
+    {
+      "epoch": 0.5167166907159242,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 10133
+    },
+    {
+      "epoch": 0.5167676841720296,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 10134
+    },
+    {
+      "epoch": 0.516818677628135,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 10135
+    },
+    {
+      "epoch": 0.5168696710842404,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 10136
+    },
+    {
+      "epoch": 0.5169206645403458,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7653,
+      "step": 10137
+    },
+    {
+      "epoch": 0.5169716579964512,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 10138
+    },
+    {
+      "epoch": 0.5170226514525565,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 10139
+    },
+    {
+      "epoch": 0.5170736449086619,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5171246383647673,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 10141
+    },
+    {
+      "epoch": 0.5171756318208728,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 10142
+    },
+    {
+      "epoch": 0.5172266252769782,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 10143
+    },
+    {
+      "epoch": 0.5172776187330835,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 10144
+    },
+    {
+      "epoch": 0.5173286121891889,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10145
+    },
+    {
+      "epoch": 0.5173796056452943,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 10146
+    },
+    {
+      "epoch": 0.5174305991013997,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 10147
+    },
+    {
+      "epoch": 0.517481592557505,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 10148
+    },
+    {
+      "epoch": 0.5175325860136105,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 10149
+    },
+    {
+      "epoch": 0.5175835794697159,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10150
+    },
+    {
+      "epoch": 0.5176345729258213,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 10151
+    },
+    {
+      "epoch": 0.5176855663819266,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 10152
+    },
+    {
+      "epoch": 0.517736559838032,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 10153
+    },
+    {
+      "epoch": 0.5177875532941374,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 10154
+    },
+    {
+      "epoch": 0.5178385467502428,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 10155
+    },
+    {
+      "epoch": 0.5178895402063483,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 10156
+    },
+    {
+      "epoch": 0.5179405336624536,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 10157
+    },
+    {
+      "epoch": 0.517991527118559,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 10158
+    },
+    {
+      "epoch": 0.5180425205746644,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 10159
+    },
+    {
+      "epoch": 0.5180935140307698,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5181445074868751,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 10161
+    },
+    {
+      "epoch": 0.5181955009429805,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 10162
+    },
+    {
+      "epoch": 0.518246494399086,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 10163
+    },
+    {
+      "epoch": 0.5182974878551914,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 10164
+    },
+    {
+      "epoch": 0.5183484813112967,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7672,
+      "step": 10165
+    },
+    {
+      "epoch": 0.5183994747674021,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 10166
+    },
+    {
+      "epoch": 0.5184504682235075,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 10167
+    },
+    {
+      "epoch": 0.5185014616796129,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 10168
+    },
+    {
+      "epoch": 0.5185524551357182,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 10169
+    },
+    {
+      "epoch": 0.5186034485918237,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5186544420479291,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 10171
+    },
+    {
+      "epoch": 0.5187054355040345,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 10172
+    },
+    {
+      "epoch": 0.5187564289601398,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 10173
+    },
+    {
+      "epoch": 0.5188074224162452,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10174
+    },
+    {
+      "epoch": 0.5188584158723506,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 10175
+    },
+    {
+      "epoch": 0.518909409328456,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 10176
+    },
+    {
+      "epoch": 0.5189604027845615,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 10177
+    },
+    {
+      "epoch": 0.5190113962406668,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 10178
+    },
+    {
+      "epoch": 0.5190623896967722,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 10179
+    },
+    {
+      "epoch": 0.5191133831528776,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 10180
+    },
+    {
+      "epoch": 0.519164376608983,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 10181
+    },
+    {
+      "epoch": 0.5192153700650883,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 10182
+    },
+    {
+      "epoch": 0.5192663635211937,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 10183
+    },
+    {
+      "epoch": 0.5193173569772992,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 10184
+    },
+    {
+      "epoch": 0.5193683504334046,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 10185
+    },
+    {
+      "epoch": 0.5194193438895099,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 10186
+    },
+    {
+      "epoch": 0.5194703373456153,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 10187
+    },
+    {
+      "epoch": 0.5195213308017207,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7746,
+      "step": 10188
+    },
+    {
+      "epoch": 0.5195723242578261,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10189
+    },
+    {
+      "epoch": 0.5196233177139314,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5196743111700369,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10191
+    },
+    {
+      "epoch": 0.5197253046261423,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 10192
+    },
+    {
+      "epoch": 0.5197762980822477,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.777,
+      "step": 10193
+    },
+    {
+      "epoch": 0.519827291538353,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 10194
+    },
+    {
+      "epoch": 0.5198782849944584,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 10195
+    },
+    {
+      "epoch": 0.5199292784505638,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 10196
+    },
+    {
+      "epoch": 0.5199802719066692,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10197
+    },
+    {
+      "epoch": 0.5200312653627747,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7738,
+      "step": 10198
+    },
+    {
+      "epoch": 0.52008225881888,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 10199
+    },
+    {
+      "epoch": 0.5201332522749854,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5201842457310908,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7829,
+      "step": 10201
+    },
+    {
+      "epoch": 0.5202352391871962,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 10202
+    },
+    {
+      "epoch": 0.5202862326433015,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10203
+    },
+    {
+      "epoch": 0.5203372260994069,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10204
+    },
+    {
+      "epoch": 0.5203882195555124,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 10205
+    },
+    {
+      "epoch": 0.5204392130116178,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10206
+    },
+    {
+      "epoch": 0.5204902064677231,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7762,
+      "step": 10207
+    },
+    {
+      "epoch": 0.5205411999238285,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 10208
+    },
+    {
+      "epoch": 0.5205921933799339,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7763,
+      "step": 10209
+    },
+    {
+      "epoch": 0.5206431868360393,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5206941802921446,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 10211
+    },
+    {
+      "epoch": 0.5207451737482501,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 10212
+    },
+    {
+      "epoch": 0.5207961672043555,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 10213
+    },
+    {
+      "epoch": 0.5208471606604609,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 10214
+    },
+    {
+      "epoch": 0.5208981541165663,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 10215
+    },
+    {
+      "epoch": 0.5209491475726716,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 10216
+    },
+    {
+      "epoch": 0.521000141028777,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 10217
+    },
+    {
+      "epoch": 0.5210511344848824,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 10218
+    },
+    {
+      "epoch": 0.5211021279409879,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 10219
+    },
+    {
+      "epoch": 0.5211531213970932,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5212041148531986,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 10221
+    },
+    {
+      "epoch": 0.521255108309304,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 10222
+    },
+    {
+      "epoch": 0.5213061017654094,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 10223
+    },
+    {
+      "epoch": 0.5213570952215147,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 10224
+    },
+    {
+      "epoch": 0.5214080886776201,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 10225
+    },
+    {
+      "epoch": 0.5214590821337256,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10226
+    },
+    {
+      "epoch": 0.521510075589831,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 10227
+    },
+    {
+      "epoch": 0.5215610690459364,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 10228
+    },
+    {
+      "epoch": 0.5216120625020417,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10229
+    },
+    {
+      "epoch": 0.5216630559581471,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5217140494142525,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 10231
+    },
+    {
+      "epoch": 0.5217650428703579,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 10232
+    },
+    {
+      "epoch": 0.5218160363264633,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 10233
+    },
+    {
+      "epoch": 0.5218670297825687,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 10234
+    },
+    {
+      "epoch": 0.5219180232386741,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 10235
+    },
+    {
+      "epoch": 0.5219690166947795,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 10236
+    },
+    {
+      "epoch": 0.5220200101508848,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 10237
+    },
+    {
+      "epoch": 0.5220710036069902,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 10238
+    },
+    {
+      "epoch": 0.5221219970630956,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 10239
+    },
+    {
+      "epoch": 0.5221729905192011,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5222239839753064,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 10241
+    },
+    {
+      "epoch": 0.5222749774314118,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 10242
+    },
+    {
+      "epoch": 0.5223259708875172,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 10243
+    },
+    {
+      "epoch": 0.5223769643436226,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 10244
+    },
+    {
+      "epoch": 0.522427957799728,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 10245
+    },
+    {
+      "epoch": 0.5224789512558333,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 10246
+    },
+    {
+      "epoch": 0.5225299447119388,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 10247
+    },
+    {
+      "epoch": 0.5225809381680442,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7798,
+      "step": 10248
+    },
+    {
+      "epoch": 0.5226319316241496,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 10249
+    },
+    {
+      "epoch": 0.5226829250802549,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5227339185363603,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 10251
+    },
+    {
+      "epoch": 0.5227849119924657,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 10252
+    },
+    {
+      "epoch": 0.5228359054485711,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10253
+    },
+    {
+      "epoch": 0.5228868989046765,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 10254
+    },
+    {
+      "epoch": 0.5229378923607819,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 10255
+    },
+    {
+      "epoch": 0.5229888858168873,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 10256
+    },
+    {
+      "epoch": 0.5230398792729927,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 10257
+    },
+    {
+      "epoch": 0.523090872729098,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 10258
+    },
+    {
+      "epoch": 0.5231418661852034,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10259
+    },
+    {
+      "epoch": 0.5231928596413088,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5232438530974143,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10261
+    },
+    {
+      "epoch": 0.5232948465535197,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 10262
+    },
+    {
+      "epoch": 0.523345840009625,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 10263
+    },
+    {
+      "epoch": 0.5233968334657304,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 10264
+    },
+    {
+      "epoch": 0.5234478269218358,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10265
+    },
+    {
+      "epoch": 0.5234988203779412,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8425,
+      "step": 10266
+    },
+    {
+      "epoch": 0.5235498138340465,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10267
+    },
+    {
+      "epoch": 0.523600807290152,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 10268
+    },
+    {
+      "epoch": 0.5236518007462574,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 10269
+    },
+    {
+      "epoch": 0.5237027942023628,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5237537876584681,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 10271
+    },
+    {
+      "epoch": 0.5238047811145735,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 10272
+    },
+    {
+      "epoch": 0.5238557745706789,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10273
+    },
+    {
+      "epoch": 0.5239067680267843,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 10274
+    },
+    {
+      "epoch": 0.5239577614828898,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 10275
+    },
+    {
+      "epoch": 0.5240087549389951,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 10276
+    },
+    {
+      "epoch": 0.5240597483951005,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7702,
+      "step": 10277
+    },
+    {
+      "epoch": 0.5241107418512059,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 10278
+    },
+    {
+      "epoch": 0.5241617353073112,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 10279
+    },
+    {
+      "epoch": 0.5242127287634166,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10280
+    },
+    {
+      "epoch": 0.524263722219522,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 10281
+    },
+    {
+      "epoch": 0.5243147156756275,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 10282
+    },
+    {
+      "epoch": 0.5243657091317329,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.774,
+      "step": 10283
+    },
+    {
+      "epoch": 0.5244167025878382,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 10284
+    },
+    {
+      "epoch": 0.5244676960439436,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 10285
+    },
+    {
+      "epoch": 0.524518689500049,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 10286
+    },
+    {
+      "epoch": 0.5245696829561544,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 10287
+    },
+    {
+      "epoch": 0.5246206764122597,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 10288
+    },
+    {
+      "epoch": 0.5246716698683652,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 10289
+    },
+    {
+      "epoch": 0.5247226633244706,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10290
+    },
+    {
+      "epoch": 0.524773656780576,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10291
+    },
+    {
+      "epoch": 0.5248246502366813,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 10292
+    },
+    {
+      "epoch": 0.5248756436927867,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 10293
+    },
+    {
+      "epoch": 0.5249266371488921,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 10294
+    },
+    {
+      "epoch": 0.5249776306049975,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 10295
+    },
+    {
+      "epoch": 0.525028624061103,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 10296
+    },
+    {
+      "epoch": 0.5250796175172083,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10297
+    },
+    {
+      "epoch": 0.5251306109733137,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 10298
+    },
+    {
+      "epoch": 0.5251816044294191,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 10299
+    },
+    {
+      "epoch": 0.5252325978855245,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5252835913416298,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 10301
+    },
+    {
+      "epoch": 0.5253345847977352,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7965,
+      "step": 10302
+    },
+    {
+      "epoch": 0.5253855782538407,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 10303
+    },
+    {
+      "epoch": 0.5254365717099461,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 10304
+    },
+    {
+      "epoch": 0.5254875651660514,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 10305
+    },
+    {
+      "epoch": 0.5255385586221568,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 10306
+    },
+    {
+      "epoch": 0.5255895520782622,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 10307
+    },
+    {
+      "epoch": 0.5256405455343676,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 10308
+    },
+    {
+      "epoch": 0.5256915389904729,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 10309
+    },
+    {
+      "epoch": 0.5257425324465784,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5257935259026838,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 10311
+    },
+    {
+      "epoch": 0.5258445193587892,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10312
+    },
+    {
+      "epoch": 0.5258955128148946,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 10313
+    },
+    {
+      "epoch": 0.5259465062709999,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 10314
+    },
+    {
+      "epoch": 0.5259974997271053,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 10315
+    },
+    {
+      "epoch": 0.5260484931832107,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7737,
+      "step": 10316
+    },
+    {
+      "epoch": 0.5260994866393162,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 10317
+    },
+    {
+      "epoch": 0.5261504800954215,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 10318
+    },
+    {
+      "epoch": 0.5262014735515269,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 10319
+    },
+    {
+      "epoch": 0.5262524670076323,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5263034604637377,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 2.7674,
+      "step": 10321
+    },
+    {
+      "epoch": 0.526354453919843,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 10322
+    },
+    {
+      "epoch": 0.5264054473759484,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 10323
+    },
+    {
+      "epoch": 0.5264564408320539,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 10324
+    },
+    {
+      "epoch": 0.5265074342881593,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 10325
+    },
+    {
+      "epoch": 0.5265584277442646,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 10326
+    },
+    {
+      "epoch": 0.52660942120037,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10327
+    },
+    {
+      "epoch": 0.5266604146564754,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 10328
+    },
+    {
+      "epoch": 0.5267114081125808,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7465,
+      "step": 10329
+    },
+    {
+      "epoch": 0.5267624015686861,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5268133950247916,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 10331
+    },
+    {
+      "epoch": 0.526864388480897,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 10332
+    },
+    {
+      "epoch": 0.5269153819370024,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 10333
+    },
+    {
+      "epoch": 0.5269663753931078,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 10334
+    },
+    {
+      "epoch": 0.5270173688492131,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 10335
+    },
+    {
+      "epoch": 0.5270683623053185,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 10336
+    },
+    {
+      "epoch": 0.5271193557614239,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10337
+    },
+    {
+      "epoch": 0.5271703492175294,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 10338
+    },
+    {
+      "epoch": 0.5272213426736347,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 10339
+    },
+    {
+      "epoch": 0.5272723361297401,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7662,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5273233295858455,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 10341
+    },
+    {
+      "epoch": 0.5273743230419509,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 10342
+    },
+    {
+      "epoch": 0.5274253164980562,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 10343
+    },
+    {
+      "epoch": 0.5274763099541616,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 10344
+    },
+    {
+      "epoch": 0.5275273034102671,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 10345
+    },
+    {
+      "epoch": 0.5275782968663725,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 10346
+    },
+    {
+      "epoch": 0.5276292903224779,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 10347
+    },
+    {
+      "epoch": 0.5276802837785832,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 10348
+    },
+    {
+      "epoch": 0.5277312772346886,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 10349
+    },
+    {
+      "epoch": 0.527782270690794,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5278332641468994,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 10351
+    },
+    {
+      "epoch": 0.5278842576030048,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 10352
+    },
+    {
+      "epoch": 0.5279352510591102,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 10353
+    },
+    {
+      "epoch": 0.5279862445152156,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 10354
+    },
+    {
+      "epoch": 0.528037237971321,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 10355
+    },
+    {
+      "epoch": 0.5280882314274263,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 10356
+    },
+    {
+      "epoch": 0.5281392248835317,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 10357
+    },
+    {
+      "epoch": 0.5281902183396371,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 10358
+    },
+    {
+      "epoch": 0.5282412117957426,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 10359
+    },
+    {
+      "epoch": 0.528292205251848,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5283431987079533,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 10361
+    },
+    {
+      "epoch": 0.5283941921640587,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 10362
+    },
+    {
+      "epoch": 0.5284451856201641,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 10363
+    },
+    {
+      "epoch": 0.5284961790762694,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 10364
+    },
+    {
+      "epoch": 0.5285471725323748,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 10365
+    },
+    {
+      "epoch": 0.5285981659884803,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 10366
+    },
+    {
+      "epoch": 0.5286491594445857,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 10367
+    },
+    {
+      "epoch": 0.5287001529006911,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8494,
+      "step": 10368
+    },
+    {
+      "epoch": 0.5287511463567964,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 10369
+    },
+    {
+      "epoch": 0.5288021398129018,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5288531332690072,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 10371
+    },
+    {
+      "epoch": 0.5289041267251126,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 10372
+    },
+    {
+      "epoch": 0.528955120181218,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 10373
+    },
+    {
+      "epoch": 0.5290061136373234,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7903,
+      "step": 10374
+    },
+    {
+      "epoch": 0.5290571070934288,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 10375
+    },
+    {
+      "epoch": 0.5291081005495342,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 10376
+    },
+    {
+      "epoch": 0.5291590940056395,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 10377
+    },
+    {
+      "epoch": 0.5292100874617449,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 10378
+    },
+    {
+      "epoch": 0.5292610809178503,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 10379
+    },
+    {
+      "epoch": 0.5293120743739558,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5293630678300612,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 10381
+    },
+    {
+      "epoch": 0.5294140612861665,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10382
+    },
+    {
+      "epoch": 0.5294650547422719,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 10383
+    },
+    {
+      "epoch": 0.5295160481983773,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 10384
+    },
+    {
+      "epoch": 0.5295670416544827,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 10385
+    },
+    {
+      "epoch": 0.529618035110588,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 10386
+    },
+    {
+      "epoch": 0.5296690285666935,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 10387
+    },
+    {
+      "epoch": 0.5297200220227989,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 10388
+    },
+    {
+      "epoch": 0.5297710154789043,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 10389
+    },
+    {
+      "epoch": 0.5298220089350096,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 10390
+    },
+    {
+      "epoch": 0.529873002391115,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 10391
+    },
+    {
+      "epoch": 0.5299239958472204,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 10392
+    },
+    {
+      "epoch": 0.5299749893033258,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 10393
+    },
+    {
+      "epoch": 0.5300259827594312,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.779,
+      "step": 10394
+    },
+    {
+      "epoch": 0.5300769762155366,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 10395
+    },
+    {
+      "epoch": 0.530127969671642,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10396
+    },
+    {
+      "epoch": 0.5301789631277474,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 10397
+    },
+    {
+      "epoch": 0.5302299565838527,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 10398
+    },
+    {
+      "epoch": 0.5302809500399581,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 10399
+    },
+    {
+      "epoch": 0.5303319434960635,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 10400
+    },
+    {
+      "epoch": 0.530382936952169,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 10401
+    },
+    {
+      "epoch": 0.5304339304082744,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 10402
+    },
+    {
+      "epoch": 0.5304849238643797,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 10403
+    },
+    {
+      "epoch": 0.5305359173204851,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 10404
+    },
+    {
+      "epoch": 0.5305869107765905,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10405
+    },
+    {
+      "epoch": 0.5306379042326959,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 10406
+    },
+    {
+      "epoch": 0.5306888976888012,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 10407
+    },
+    {
+      "epoch": 0.5307398911449067,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 10408
+    },
+    {
+      "epoch": 0.5307908846010121,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 10409
+    },
+    {
+      "epoch": 0.5308418780571175,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5308928715132228,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 10411
+    },
+    {
+      "epoch": 0.5309438649693282,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 10412
+    },
+    {
+      "epoch": 0.5309948584254336,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 10413
+    },
+    {
+      "epoch": 0.531045851881539,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 10414
+    },
+    {
+      "epoch": 0.5310968453376445,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 10415
+    },
+    {
+      "epoch": 0.5311478387937498,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 10416
+    },
+    {
+      "epoch": 0.5311988322498552,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 10417
+    },
+    {
+      "epoch": 0.5312498257059606,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 10418
+    },
+    {
+      "epoch": 0.531300819162066,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 10419
+    },
+    {
+      "epoch": 0.5313518126181713,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5314028060742767,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 10421
+    },
+    {
+      "epoch": 0.5314537995303822,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 10422
+    },
+    {
+      "epoch": 0.5315047929864876,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 10423
+    },
+    {
+      "epoch": 0.5315557864425929,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 10424
+    },
+    {
+      "epoch": 0.5316067798986983,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10425
+    },
+    {
+      "epoch": 0.5316577733548037,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 10426
+    },
+    {
+      "epoch": 0.5317087668109091,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 10427
+    },
+    {
+      "epoch": 0.5317597602670144,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 10428
+    },
+    {
+      "epoch": 0.5318107537231199,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 10429
+    },
+    {
+      "epoch": 0.5318617471792253,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5319127406353307,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 10431
+    },
+    {
+      "epoch": 0.531963734091436,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7777,
+      "step": 10432
+    },
+    {
+      "epoch": 0.5320147275475414,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10433
+    },
+    {
+      "epoch": 0.5320657210036468,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 10434
+    },
+    {
+      "epoch": 0.5321167144597522,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 10435
+    },
+    {
+      "epoch": 0.5321677079158577,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10436
+    },
+    {
+      "epoch": 0.532218701371963,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 10437
+    },
+    {
+      "epoch": 0.5322696948280684,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8441,
+      "step": 10438
+    },
+    {
+      "epoch": 0.5323206882841738,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 10439
+    },
+    {
+      "epoch": 0.5323716817402792,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5324226751963845,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 10441
+    },
+    {
+      "epoch": 0.5324736686524899,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 10442
+    },
+    {
+      "epoch": 0.5325246621085954,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 10443
+    },
+    {
+      "epoch": 0.5325756555647008,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 10444
+    },
+    {
+      "epoch": 0.5326266490208061,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10445
+    },
+    {
+      "epoch": 0.5326776424769115,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 10446
+    },
+    {
+      "epoch": 0.5327286359330169,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10447
+    },
+    {
+      "epoch": 0.5327796293891223,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10448
+    },
+    {
+      "epoch": 0.5328306228452276,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7727,
+      "step": 10449
+    },
+    {
+      "epoch": 0.5328816163013331,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5329326097574385,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 10451
+    },
+    {
+      "epoch": 0.5329836032135439,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7866,
+      "step": 10452
+    },
+    {
+      "epoch": 0.5330345966696493,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 10453
+    },
+    {
+      "epoch": 0.5330855901257546,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 10454
+    },
+    {
+      "epoch": 0.53313658358186,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 10455
+    },
+    {
+      "epoch": 0.5331875770379654,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10456
+    },
+    {
+      "epoch": 0.5332385704940709,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10457
+    },
+    {
+      "epoch": 0.5332895639501762,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 10458
+    },
+    {
+      "epoch": 0.5333405574062816,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 10459
+    },
+    {
+      "epoch": 0.533391550862387,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5334425443184924,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 10461
+    },
+    {
+      "epoch": 0.5334935377745977,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 10462
+    },
+    {
+      "epoch": 0.5335445312307031,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 10463
+    },
+    {
+      "epoch": 0.5335955246868086,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7614,
+      "step": 10464
+    },
+    {
+      "epoch": 0.533646518142914,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 10465
+    },
+    {
+      "epoch": 0.5336975115990193,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 10466
+    },
+    {
+      "epoch": 0.5337485050551247,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 10467
+    },
+    {
+      "epoch": 0.5337994985112301,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 10468
+    },
+    {
+      "epoch": 0.5338504919673355,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 10469
+    },
+    {
+      "epoch": 0.5339014854234408,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5339524788795463,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 10471
+    },
+    {
+      "epoch": 0.5340034723356517,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 10472
+    },
+    {
+      "epoch": 0.5340544657917571,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 10473
+    },
+    {
+      "epoch": 0.5341054592478625,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 10474
+    },
+    {
+      "epoch": 0.5341564527039678,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 10475
+    },
+    {
+      "epoch": 0.5342074461600732,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 10476
+    },
+    {
+      "epoch": 0.5342584396161786,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 10477
+    },
+    {
+      "epoch": 0.5343094330722841,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 10478
+    },
+    {
+      "epoch": 0.5343604265283894,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 10479
+    },
+    {
+      "epoch": 0.5344114199844948,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5344624134406002,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 10481
+    },
+    {
+      "epoch": 0.5345134068967056,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 10482
+    },
+    {
+      "epoch": 0.5345644003528109,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10483
+    },
+    {
+      "epoch": 0.5346153938089163,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 10484
+    },
+    {
+      "epoch": 0.5346663872650218,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 10485
+    },
+    {
+      "epoch": 0.5347173807211272,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 10486
+    },
+    {
+      "epoch": 0.5347683741772326,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 10487
+    },
+    {
+      "epoch": 0.5348193676333379,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10488
+    },
+    {
+      "epoch": 0.5348703610894433,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 10489
+    },
+    {
+      "epoch": 0.5349213545455487,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 10490
+    },
+    {
+      "epoch": 0.534972348001654,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 10491
+    },
+    {
+      "epoch": 0.5350233414577595,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 10492
+    },
+    {
+      "epoch": 0.5350743349138649,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 10493
+    },
+    {
+      "epoch": 0.5351253283699703,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7722,
+      "step": 10494
+    },
+    {
+      "epoch": 0.5351763218260757,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 10495
+    },
+    {
+      "epoch": 0.535227315282181,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 10496
+    },
+    {
+      "epoch": 0.5352783087382864,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 10497
+    },
+    {
+      "epoch": 0.5353293021943918,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10498
+    },
+    {
+      "epoch": 0.5353802956504973,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 10499
+    },
+    {
+      "epoch": 0.5354312891066026,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 10500
+    },
+    {
+      "epoch": 0.535482282562708,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 10501
+    },
+    {
+      "epoch": 0.5355332760188134,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7753,
+      "step": 10502
+    },
+    {
+      "epoch": 0.5355842694749188,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 10503
+    },
+    {
+      "epoch": 0.5356352629310241,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 10504
+    },
+    {
+      "epoch": 0.5356862563871295,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10505
+    },
+    {
+      "epoch": 0.535737249843235,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 10506
+    },
+    {
+      "epoch": 0.5357882432993404,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 10507
+    },
+    {
+      "epoch": 0.5358392367554458,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 10508
+    },
+    {
+      "epoch": 0.5358902302115511,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 10509
+    },
+    {
+      "epoch": 0.5359412236676565,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5359922171237619,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 10511
+    },
+    {
+      "epoch": 0.5360432105798673,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 10512
+    },
+    {
+      "epoch": 0.5360942040359727,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 10513
+    },
+    {
+      "epoch": 0.5361451974920781,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 10514
+    },
+    {
+      "epoch": 0.5361961909481835,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 10515
+    },
+    {
+      "epoch": 0.5362471844042889,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10516
+    },
+    {
+      "epoch": 0.5362981778603942,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10517
+    },
+    {
+      "epoch": 0.5363491713164996,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 10518
+    },
+    {
+      "epoch": 0.536400164772605,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7799,
+      "step": 10519
+    },
+    {
+      "epoch": 0.5364511582287105,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5365021516848159,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 10521
+    },
+    {
+      "epoch": 0.5365531451409212,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 10522
+    },
+    {
+      "epoch": 0.5366041385970266,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 10523
+    },
+    {
+      "epoch": 0.536655132053132,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 10524
+    },
+    {
+      "epoch": 0.5367061255092374,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 10525
+    },
+    {
+      "epoch": 0.5367571189653427,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 10526
+    },
+    {
+      "epoch": 0.5368081124214482,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 10527
+    },
+    {
+      "epoch": 0.5368591058775536,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 10528
+    },
+    {
+      "epoch": 0.536910099333659,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10529
+    },
+    {
+      "epoch": 0.5369610927897643,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5370120862458697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 10531
+    },
+    {
+      "epoch": 0.5370630797019751,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 10532
+    },
+    {
+      "epoch": 0.5371140731580805,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7835,
+      "step": 10533
+    },
+    {
+      "epoch": 0.537165066614186,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 10534
+    },
+    {
+      "epoch": 0.5372160600702913,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7761,
+      "step": 10535
+    },
+    {
+      "epoch": 0.5372670535263967,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10536
+    },
+    {
+      "epoch": 0.5373180469825021,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 10537
+    },
+    {
+      "epoch": 0.5373690404386074,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 10538
+    },
+    {
+      "epoch": 0.5374200338947128,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 10539
+    },
+    {
+      "epoch": 0.5374710273508182,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5375220208069237,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 10541
+    },
+    {
+      "epoch": 0.5375730142630291,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 10542
+    },
+    {
+      "epoch": 0.5376240077191344,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 10543
+    },
+    {
+      "epoch": 0.5376750011752398,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 10544
+    },
+    {
+      "epoch": 0.5377259946313452,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 10545
+    },
+    {
+      "epoch": 0.5377769880874506,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 10546
+    },
+    {
+      "epoch": 0.5378279815435559,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 10547
+    },
+    {
+      "epoch": 0.5378789749996614,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 10548
+    },
+    {
+      "epoch": 0.5379299684557668,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 10549
+    },
+    {
+      "epoch": 0.5379809619118722,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5380319553679775,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 10551
+    },
+    {
+      "epoch": 0.5380829488240829,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10552
+    },
+    {
+      "epoch": 0.5381339422801883,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 10553
+    },
+    {
+      "epoch": 0.5381849357362937,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 10554
+    },
+    {
+      "epoch": 0.5382359291923992,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 10555
+    },
+    {
+      "epoch": 0.5382869226485045,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 10556
+    },
+    {
+      "epoch": 0.5383379161046099,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 10557
+    },
+    {
+      "epoch": 0.5383889095607153,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 10558
+    },
+    {
+      "epoch": 0.5384399030168207,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7762,
+      "step": 10559
+    },
+    {
+      "epoch": 0.538490896472926,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5385418899290314,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 10561
+    },
+    {
+      "epoch": 0.5385928833851369,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 10562
+    },
+    {
+      "epoch": 0.5386438768412423,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 10563
+    },
+    {
+      "epoch": 0.5386948702973476,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 10564
+    },
+    {
+      "epoch": 0.538745863753453,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 10565
+    },
+    {
+      "epoch": 0.5387968572095584,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 10566
+    },
+    {
+      "epoch": 0.5388478506656638,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 10567
+    },
+    {
+      "epoch": 0.5388988441217691,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 10568
+    },
+    {
+      "epoch": 0.5389498375778746,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10569
+    },
+    {
+      "epoch": 0.53900083103398,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5390518244900854,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 10571
+    },
+    {
+      "epoch": 0.5391028179461907,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 10572
+    },
+    {
+      "epoch": 0.5391538114022961,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 10573
+    },
+    {
+      "epoch": 0.5392048048584015,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 10574
+    },
+    {
+      "epoch": 0.5392557983145069,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 10575
+    },
+    {
+      "epoch": 0.5393067917706124,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 10576
+    },
+    {
+      "epoch": 0.5393577852267177,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 10577
+    },
+    {
+      "epoch": 0.5394087786828231,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 10578
+    },
+    {
+      "epoch": 0.5394597721389285,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 10579
+    },
+    {
+      "epoch": 0.5395107655950339,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5395617590511392,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 10581
+    },
+    {
+      "epoch": 0.5396127525072446,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 10582
+    },
+    {
+      "epoch": 0.5396637459633501,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 10583
+    },
+    {
+      "epoch": 0.5397147394194555,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10584
+    },
+    {
+      "epoch": 0.5397657328755608,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 10585
+    },
+    {
+      "epoch": 0.5398167263316662,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7648,
+      "step": 10586
+    },
+    {
+      "epoch": 0.5398677197877716,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 10587
+    },
+    {
+      "epoch": 0.539918713243877,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7737,
+      "step": 10588
+    },
+    {
+      "epoch": 0.5399697066999823,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 10589
+    },
+    {
+      "epoch": 0.5400207001560878,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5400716936121932,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 10591
+    },
+    {
+      "epoch": 0.5401226870682986,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 10592
+    },
+    {
+      "epoch": 0.540173680524404,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 10593
+    },
+    {
+      "epoch": 0.5402246739805093,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 10594
+    },
+    {
+      "epoch": 0.5402756674366147,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10595
+    },
+    {
+      "epoch": 0.5403266608927201,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 10596
+    },
+    {
+      "epoch": 0.5403776543488256,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 10597
+    },
+    {
+      "epoch": 0.5404286478049309,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 10598
+    },
+    {
+      "epoch": 0.5404796412610363,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 10599
+    },
+    {
+      "epoch": 0.5405306347171417,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5405816281732471,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 10601
+    },
+    {
+      "epoch": 0.5406326216293524,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 10602
+    },
+    {
+      "epoch": 0.5406836150854578,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 10603
+    },
+    {
+      "epoch": 0.5407346085415633,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 10604
+    },
+    {
+      "epoch": 0.5407856019976687,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 10605
+    },
+    {
+      "epoch": 0.540836595453774,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 10606
+    },
+    {
+      "epoch": 0.5408875889098794,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 10607
+    },
+    {
+      "epoch": 0.5409385823659848,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 10608
+    },
+    {
+      "epoch": 0.5409895758220902,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 10609
+    },
+    {
+      "epoch": 0.5410405692781955,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 10610
+    },
+    {
+      "epoch": 0.541091562734301,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 10611
+    },
+    {
+      "epoch": 0.5411425561904064,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 10612
+    },
+    {
+      "epoch": 0.5411935496465118,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 10613
+    },
+    {
+      "epoch": 0.5412445431026172,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 10614
+    },
+    {
+      "epoch": 0.5412955365587225,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 10615
+    },
+    {
+      "epoch": 0.5413465300148279,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 10616
+    },
+    {
+      "epoch": 0.5413975234709333,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 10617
+    },
+    {
+      "epoch": 0.5414485169270388,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 10618
+    },
+    {
+      "epoch": 0.5414995103831441,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 10619
+    },
+    {
+      "epoch": 0.5415505038392495,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5416014972953549,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 10621
+    },
+    {
+      "epoch": 0.5416524907514603,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8372,
+      "step": 10622
+    },
+    {
+      "epoch": 0.5417034842075656,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 10623
+    },
+    {
+      "epoch": 0.541754477663671,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 10624
+    },
+    {
+      "epoch": 0.5418054711197765,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 10625
+    },
+    {
+      "epoch": 0.5418564645758819,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 10626
+    },
+    {
+      "epoch": 0.5419074580319873,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 10627
+    },
+    {
+      "epoch": 0.5419584514880926,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 10628
+    },
+    {
+      "epoch": 0.542009444944198,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 10629
+    },
+    {
+      "epoch": 0.5420604384003034,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5421114318564088,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 10631
+    },
+    {
+      "epoch": 0.5421624253125142,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 10632
+    },
+    {
+      "epoch": 0.5422134187686196,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 10633
+    },
+    {
+      "epoch": 0.542264412224725,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 10634
+    },
+    {
+      "epoch": 0.5423154056808304,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 10635
+    },
+    {
+      "epoch": 0.5423663991369357,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 10636
+    },
+    {
+      "epoch": 0.5424173925930411,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 10637
+    },
+    {
+      "epoch": 0.5424683860491465,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 10638
+    },
+    {
+      "epoch": 0.542519379505252,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 10639
+    },
+    {
+      "epoch": 0.5425703729613574,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5426213664174627,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 10641
+    },
+    {
+      "epoch": 0.5426723598735681,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 10642
+    },
+    {
+      "epoch": 0.5427233533296735,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 10643
+    },
+    {
+      "epoch": 0.5427743467857788,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10644
+    },
+    {
+      "epoch": 0.5428253402418842,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 10645
+    },
+    {
+      "epoch": 0.5428763336979897,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 10646
+    },
+    {
+      "epoch": 0.5429273271540951,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 10647
+    },
+    {
+      "epoch": 0.5429783206102005,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 10648
+    },
+    {
+      "epoch": 0.5430293140663058,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 10649
+    },
+    {
+      "epoch": 0.5430803075224112,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5431313009785166,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 10651
+    },
+    {
+      "epoch": 0.543182294434622,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 10652
+    },
+    {
+      "epoch": 0.5432332878907274,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 10653
+    },
+    {
+      "epoch": 0.5432842813468328,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 10654
+    },
+    {
+      "epoch": 0.5433352748029382,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 10655
+    },
+    {
+      "epoch": 0.5433862682590436,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 10656
+    },
+    {
+      "epoch": 0.543437261715149,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 10657
+    },
+    {
+      "epoch": 0.5434882551712543,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 10658
+    },
+    {
+      "epoch": 0.5435392486273597,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 10659
+    },
+    {
+      "epoch": 0.5435902420834652,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5436412355395706,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 10661
+    },
+    {
+      "epoch": 0.5436922289956759,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7573,
+      "step": 10662
+    },
+    {
+      "epoch": 0.5437432224517813,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 10663
+    },
+    {
+      "epoch": 0.5437942159078867,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 10664
+    },
+    {
+      "epoch": 0.543845209363992,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 10665
+    },
+    {
+      "epoch": 0.5438962028200974,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10666
+    },
+    {
+      "epoch": 0.5439471962762029,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7829,
+      "step": 10667
+    },
+    {
+      "epoch": 0.5439981897323083,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 10668
+    },
+    {
+      "epoch": 0.5440491831884137,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 10669
+    },
+    {
+      "epoch": 0.544100176644519,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8455,
+      "step": 10670
+    },
+    {
+      "epoch": 0.5441511701006244,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 10671
+    },
+    {
+      "epoch": 0.5442021635567298,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 10672
+    },
+    {
+      "epoch": 0.5442531570128352,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 10673
+    },
+    {
+      "epoch": 0.5443041504689407,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 10674
+    },
+    {
+      "epoch": 0.544355143925046,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 10675
+    },
+    {
+      "epoch": 0.5444061373811514,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7845,
+      "step": 10676
+    },
+    {
+      "epoch": 0.5444571308372568,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10677
+    },
+    {
+      "epoch": 0.5445081242933622,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 10678
+    },
+    {
+      "epoch": 0.5445591177494675,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 10679
+    },
+    {
+      "epoch": 0.5446101112055729,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5446611046616784,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10681
+    },
+    {
+      "epoch": 0.5447120981177838,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 10682
+    },
+    {
+      "epoch": 0.5447630915738891,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 10683
+    },
+    {
+      "epoch": 0.5448140850299945,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 10684
+    },
+    {
+      "epoch": 0.5448650784860999,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 10685
+    },
+    {
+      "epoch": 0.5449160719422053,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 10686
+    },
+    {
+      "epoch": 0.5449670653983106,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 10687
+    },
+    {
+      "epoch": 0.5450180588544161,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 10688
+    },
+    {
+      "epoch": 0.5450690523105215,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8422,
+      "step": 10689
+    },
+    {
+      "epoch": 0.5451200457666269,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5451710392227322,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 10691
+    },
+    {
+      "epoch": 0.5452220326788376,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10692
+    },
+    {
+      "epoch": 0.545273026134943,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 10693
+    },
+    {
+      "epoch": 0.5453240195910484,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 10694
+    },
+    {
+      "epoch": 0.5453750130471539,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10695
+    },
+    {
+      "epoch": 0.5454260065032592,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 10696
+    },
+    {
+      "epoch": 0.5454769999593646,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 10697
+    },
+    {
+      "epoch": 0.54552799341547,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7801,
+      "step": 10698
+    },
+    {
+      "epoch": 0.5455789868715754,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7788,
+      "step": 10699
+    },
+    {
+      "epoch": 0.5456299803276807,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5456809737837861,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 10701
+    },
+    {
+      "epoch": 0.5457319672398916,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 10702
+    },
+    {
+      "epoch": 0.545782960695997,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10703
+    },
+    {
+      "epoch": 0.5458339541521023,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 10704
+    },
+    {
+      "epoch": 0.5458849476082077,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 10705
+    },
+    {
+      "epoch": 0.5459359410643131,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 10706
+    },
+    {
+      "epoch": 0.5459869345204185,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 10707
+    },
+    {
+      "epoch": 0.5460379279765238,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 10708
+    },
+    {
+      "epoch": 0.5460889214326293,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 10709
+    },
+    {
+      "epoch": 0.5461399148887347,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5461909083448401,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 10711
+    },
+    {
+      "epoch": 0.5462419018009455,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 10712
+    },
+    {
+      "epoch": 0.5462928952570508,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7701,
+      "step": 10713
+    },
+    {
+      "epoch": 0.5463438887131562,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.782,
+      "step": 10714
+    },
+    {
+      "epoch": 0.5463948821692616,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 10715
+    },
+    {
+      "epoch": 0.5464458756253671,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 10716
+    },
+    {
+      "epoch": 0.5464968690814724,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 10717
+    },
+    {
+      "epoch": 0.5465478625375778,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 10718
+    },
+    {
+      "epoch": 0.5465988559936832,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 10719
+    },
+    {
+      "epoch": 0.5466498494497886,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5467008429058939,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 10721
+    },
+    {
+      "epoch": 0.5467518363619993,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10722
+    },
+    {
+      "epoch": 0.5468028298181048,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 10723
+    },
+    {
+      "epoch": 0.5468538232742102,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 10724
+    },
+    {
+      "epoch": 0.5469048167303155,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 10725
+    },
+    {
+      "epoch": 0.5469558101864209,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 10726
+    },
+    {
+      "epoch": 0.5470068036425263,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 10727
+    },
+    {
+      "epoch": 0.5470577970986317,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 10728
+    },
+    {
+      "epoch": 0.547108790554737,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 10729
+    },
+    {
+      "epoch": 0.5471597840108425,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5472107774669479,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 10731
+    },
+    {
+      "epoch": 0.5472617709230533,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 10732
+    },
+    {
+      "epoch": 0.5473127643791587,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10733
+    },
+    {
+      "epoch": 0.547363757835264,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 10734
+    },
+    {
+      "epoch": 0.5474147512913694,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 10735
+    },
+    {
+      "epoch": 0.5474657447474748,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 10736
+    },
+    {
+      "epoch": 0.5475167382035803,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7442,
+      "step": 10737
+    },
+    {
+      "epoch": 0.5475677316596856,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 10738
+    },
+    {
+      "epoch": 0.547618725115791,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 10739
+    },
+    {
+      "epoch": 0.5476697185718964,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5477207120280018,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 10741
+    },
+    {
+      "epoch": 0.5477717054841071,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 10742
+    },
+    {
+      "epoch": 0.5478226989402125,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 10743
+    },
+    {
+      "epoch": 0.547873692396318,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 10744
+    },
+    {
+      "epoch": 0.5479246858524234,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 10745
+    },
+    {
+      "epoch": 0.5479756793085288,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 10746
+    },
+    {
+      "epoch": 0.5480266727646341,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 10747
+    },
+    {
+      "epoch": 0.5480776662207395,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 10748
+    },
+    {
+      "epoch": 0.5481286596768449,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 10749
+    },
+    {
+      "epoch": 0.5481796531329503,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8675,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5482306465890557,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 10751
+    },
+    {
+      "epoch": 0.5482816400451611,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 10752
+    },
+    {
+      "epoch": 0.5483326335012665,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 10753
+    },
+    {
+      "epoch": 0.5483836269573719,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 10754
+    },
+    {
+      "epoch": 0.5484346204134772,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 10755
+    },
+    {
+      "epoch": 0.5484856138695826,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7662,
+      "step": 10756
+    },
+    {
+      "epoch": 0.548536607325688,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 10757
+    },
+    {
+      "epoch": 0.5485876007817935,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 10758
+    },
+    {
+      "epoch": 0.5486385942378988,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 10759
+    },
+    {
+      "epoch": 0.5486895876940042,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5487405811501096,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 10761
+    },
+    {
+      "epoch": 0.548791574606215,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 10762
+    },
+    {
+      "epoch": 0.5488425680623203,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 10763
+    },
+    {
+      "epoch": 0.5488935615184257,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 10764
+    },
+    {
+      "epoch": 0.5489445549745312,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 10765
+    },
+    {
+      "epoch": 0.5489955484306366,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10766
+    },
+    {
+      "epoch": 0.549046541886742,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 10767
+    },
+    {
+      "epoch": 0.5490975353428473,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 10768
+    },
+    {
+      "epoch": 0.5491485287989527,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 10769
+    },
+    {
+      "epoch": 0.5491995222550581,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5492505157111635,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 10771
+    },
+    {
+      "epoch": 0.5493015091672689,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 10772
+    },
+    {
+      "epoch": 0.5493525026233743,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 10773
+    },
+    {
+      "epoch": 0.5494034960794797,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7872,
+      "step": 10774
+    },
+    {
+      "epoch": 0.5494544895355851,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 10775
+    },
+    {
+      "epoch": 0.5495054829916904,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 10776
+    },
+    {
+      "epoch": 0.5495564764477958,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7662,
+      "step": 10777
+    },
+    {
+      "epoch": 0.5496074699039012,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 10778
+    },
+    {
+      "epoch": 0.5496584633600067,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 10779
+    },
+    {
+      "epoch": 0.549709456816112,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5497604502722174,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 10781
+    },
+    {
+      "epoch": 0.5498114437283228,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 10782
+    },
+    {
+      "epoch": 0.5498624371844282,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 10783
+    },
+    {
+      "epoch": 0.5499134306405336,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 10784
+    },
+    {
+      "epoch": 0.5499644240966389,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 10785
+    },
+    {
+      "epoch": 0.5500154175527444,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 10786
+    },
+    {
+      "epoch": 0.5500664110088498,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 10787
+    },
+    {
+      "epoch": 0.5501174044649552,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 10788
+    },
+    {
+      "epoch": 0.5501683979210605,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 10789
+    },
+    {
+      "epoch": 0.5502193913771659,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5502703848332713,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 10791
+    },
+    {
+      "epoch": 0.5503213782893767,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 10792
+    },
+    {
+      "epoch": 0.5503723717454821,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 10793
+    },
+    {
+      "epoch": 0.5504233652015875,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7608,
+      "step": 10794
+    },
+    {
+      "epoch": 0.5504743586576929,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 10795
+    },
+    {
+      "epoch": 0.5505253521137983,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10796
+    },
+    {
+      "epoch": 0.5505763455699036,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 10797
+    },
+    {
+      "epoch": 0.550627339026009,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 10798
+    },
+    {
+      "epoch": 0.5506783324821144,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 10799
+    },
+    {
+      "epoch": 0.5507293259382199,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5507803193943253,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 10801
+    },
+    {
+      "epoch": 0.5508313128504306,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 10802
+    },
+    {
+      "epoch": 0.550882306306536,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 10803
+    },
+    {
+      "epoch": 0.5509332997626414,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10804
+    },
+    {
+      "epoch": 0.5509842932187468,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10805
+    },
+    {
+      "epoch": 0.5510352866748521,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 10806
+    },
+    {
+      "epoch": 0.5510862801309576,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 10807
+    },
+    {
+      "epoch": 0.551137273587063,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 10808
+    },
+    {
+      "epoch": 0.5511882670431684,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8283,
+      "step": 10809
+    },
+    {
+      "epoch": 0.5512392604992737,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5512902539553791,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 10811
+    },
+    {
+      "epoch": 0.5513412474114845,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 10812
+    },
+    {
+      "epoch": 0.5513922408675899,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 10813
+    },
+    {
+      "epoch": 0.5514432343236954,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 10814
+    },
+    {
+      "epoch": 0.5514942277798007,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 10815
+    },
+    {
+      "epoch": 0.5515452212359061,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 10816
+    },
+    {
+      "epoch": 0.5515962146920115,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 10817
+    },
+    {
+      "epoch": 0.5516472081481169,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 10818
+    },
+    {
+      "epoch": 0.5516982016042222,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 10819
+    },
+    {
+      "epoch": 0.5517491950603276,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7655,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5518001885164331,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 10821
+    },
+    {
+      "epoch": 0.5518511819725385,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 10822
+    },
+    {
+      "epoch": 0.5519021754286438,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 10823
+    },
+    {
+      "epoch": 0.5519531688847492,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 10824
+    },
+    {
+      "epoch": 0.5520041623408546,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10825
+    },
+    {
+      "epoch": 0.55205515579696,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 10826
+    },
+    {
+      "epoch": 0.5521061492530653,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7903,
+      "step": 10827
+    },
+    {
+      "epoch": 0.5521571427091708,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 10828
+    },
+    {
+      "epoch": 0.5522081361652762,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 10829
+    },
+    {
+      "epoch": 0.5522591296213816,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 10830
+    },
+    {
+      "epoch": 0.552310123077487,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 10831
+    },
+    {
+      "epoch": 0.5523611165335923,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 10832
+    },
+    {
+      "epoch": 0.5524121099896977,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 10833
+    },
+    {
+      "epoch": 0.5524631034458031,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7772,
+      "step": 10834
+    },
+    {
+      "epoch": 0.5525140969019086,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10835
+    },
+    {
+      "epoch": 0.5525650903580139,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 10836
+    },
+    {
+      "epoch": 0.5526160838141193,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 10837
+    },
+    {
+      "epoch": 0.5526670772702247,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 10838
+    },
+    {
+      "epoch": 0.5527180707263301,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10839
+    },
+    {
+      "epoch": 0.5527690641824354,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5528200576385408,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 10841
+    },
+    {
+      "epoch": 0.5528710510946463,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 10842
+    },
+    {
+      "epoch": 0.5529220445507517,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 10843
+    },
+    {
+      "epoch": 0.552973038006857,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 10844
+    },
+    {
+      "epoch": 0.5530240314629624,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10845
+    },
+    {
+      "epoch": 0.5530750249190678,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 10846
+    },
+    {
+      "epoch": 0.5531260183751732,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 10847
+    },
+    {
+      "epoch": 0.5531770118312785,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 10848
+    },
+    {
+      "epoch": 0.553228005287384,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7835,
+      "step": 10849
+    },
+    {
+      "epoch": 0.5532789987434894,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5533299921995948,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10851
+    },
+    {
+      "epoch": 0.5533809856557002,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 10852
+    },
+    {
+      "epoch": 0.5534319791118055,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 10853
+    },
+    {
+      "epoch": 0.5534829725679109,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 10854
+    },
+    {
+      "epoch": 0.5535339660240163,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 10855
+    },
+    {
+      "epoch": 0.5535849594801218,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 10856
+    },
+    {
+      "epoch": 0.5536359529362271,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 10857
+    },
+    {
+      "epoch": 0.5536869463923325,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 10858
+    },
+    {
+      "epoch": 0.5537379398484379,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 10859
+    },
+    {
+      "epoch": 0.5537889333045433,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5538399267606486,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 10861
+    },
+    {
+      "epoch": 0.553890920216754,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 10862
+    },
+    {
+      "epoch": 0.5539419136728595,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 10863
+    },
+    {
+      "epoch": 0.5539929071289649,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 10864
+    },
+    {
+      "epoch": 0.5540439005850702,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7758,
+      "step": 10865
+    },
+    {
+      "epoch": 0.5540948940411756,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 10866
+    },
+    {
+      "epoch": 0.554145887497281,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 10867
+    },
+    {
+      "epoch": 0.5541968809533864,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10868
+    },
+    {
+      "epoch": 0.5542478744094917,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7754,
+      "step": 10869
+    },
+    {
+      "epoch": 0.5542988678655972,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5543498613217026,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 10871
+    },
+    {
+      "epoch": 0.554400854777808,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 10872
+    },
+    {
+      "epoch": 0.5544518482339134,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 10873
+    },
+    {
+      "epoch": 0.5545028416900187,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 10874
+    },
+    {
+      "epoch": 0.5545538351461241,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 10875
+    },
+    {
+      "epoch": 0.5546048286022295,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 10876
+    },
+    {
+      "epoch": 0.554655822058335,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 10877
+    },
+    {
+      "epoch": 0.5547068155144403,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 10878
+    },
+    {
+      "epoch": 0.5547578089705457,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 10879
+    },
+    {
+      "epoch": 0.5548088024266511,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5548597958827565,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 10881
+    },
+    {
+      "epoch": 0.5549107893388618,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 10882
+    },
+    {
+      "epoch": 0.5549617827949672,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 10883
+    },
+    {
+      "epoch": 0.5550127762510727,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 10884
+    },
+    {
+      "epoch": 0.5550637697071781,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 10885
+    },
+    {
+      "epoch": 0.5551147631632835,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 10886
+    },
+    {
+      "epoch": 0.5551657566193888,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 10887
+    },
+    {
+      "epoch": 0.5552167500754942,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 10888
+    },
+    {
+      "epoch": 0.5552677435315996,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 10889
+    },
+    {
+      "epoch": 0.555318736987705,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5553697304438104,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 10891
+    },
+    {
+      "epoch": 0.5554207238999158,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 10892
+    },
+    {
+      "epoch": 0.5554717173560212,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10893
+    },
+    {
+      "epoch": 0.5555227108121266,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 10894
+    },
+    {
+      "epoch": 0.5555737042682319,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 10895
+    },
+    {
+      "epoch": 0.5556246977243373,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 10896
+    },
+    {
+      "epoch": 0.5556756911804427,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 10897
+    },
+    {
+      "epoch": 0.5557266846365482,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 10898
+    },
+    {
+      "epoch": 0.5557776780926535,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 10899
+    },
+    {
+      "epoch": 0.5558286715487589,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5558796650048643,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 10901
+    },
+    {
+      "epoch": 0.5559306584609697,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8449,
+      "step": 10902
+    },
+    {
+      "epoch": 0.555981651917075,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 10903
+    },
+    {
+      "epoch": 0.5560326453731804,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 10904
+    },
+    {
+      "epoch": 0.5560836388292859,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 10905
+    },
+    {
+      "epoch": 0.5561346322853913,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 10906
+    },
+    {
+      "epoch": 0.5561856257414967,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 10907
+    },
+    {
+      "epoch": 0.556236619197602,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10908
+    },
+    {
+      "epoch": 0.5562876126537074,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 10909
+    },
+    {
+      "epoch": 0.5563386061098128,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5563895995659182,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 10911
+    },
+    {
+      "epoch": 0.5564405930220236,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 10912
+    },
+    {
+      "epoch": 0.556491586478129,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7637,
+      "step": 10913
+    },
+    {
+      "epoch": 0.5565425799342344,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 10914
+    },
+    {
+      "epoch": 0.5565935733903398,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 10915
+    },
+    {
+      "epoch": 0.5566445668464451,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 10916
+    },
+    {
+      "epoch": 0.5566955603025505,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.77,
+      "step": 10917
+    },
+    {
+      "epoch": 0.5567465537586559,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 10918
+    },
+    {
+      "epoch": 0.5567975472147614,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 10919
+    },
+    {
+      "epoch": 0.5568485406708668,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5568995341269721,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7603,
+      "step": 10921
+    },
+    {
+      "epoch": 0.5569505275830775,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 10922
+    },
+    {
+      "epoch": 0.5570015210391829,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7742,
+      "step": 10923
+    },
+    {
+      "epoch": 0.5570525144952883,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 10924
+    },
+    {
+      "epoch": 0.5571035079513936,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 10925
+    },
+    {
+      "epoch": 0.5571545014074991,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 10926
+    },
+    {
+      "epoch": 0.5572054948636045,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 10927
+    },
+    {
+      "epoch": 0.5572564883197099,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7505,
+      "step": 10928
+    },
+    {
+      "epoch": 0.5573074817758152,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10929
+    },
+    {
+      "epoch": 0.5573584752319206,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 10930
+    },
+    {
+      "epoch": 0.557409468688026,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 10931
+    },
+    {
+      "epoch": 0.5574604621441314,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 10932
+    },
+    {
+      "epoch": 0.5575114556002368,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 10933
+    },
+    {
+      "epoch": 0.5575624490563422,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 10934
+    },
+    {
+      "epoch": 0.5576134425124476,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 10935
+    },
+    {
+      "epoch": 0.557664435968553,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 10936
+    },
+    {
+      "epoch": 0.5577154294246583,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 10937
+    },
+    {
+      "epoch": 0.5577664228807637,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 10938
+    },
+    {
+      "epoch": 0.5578174163368691,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 10939
+    },
+    {
+      "epoch": 0.5578684097929746,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 10940
+    },
+    {
+      "epoch": 0.55791940324908,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 10941
+    },
+    {
+      "epoch": 0.5579703967051853,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 10942
+    },
+    {
+      "epoch": 0.5580213901612907,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 10943
+    },
+    {
+      "epoch": 0.5580723836173961,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 10944
+    },
+    {
+      "epoch": 0.5581233770735015,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5581743705296068,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 10946
+    },
+    {
+      "epoch": 0.5582253639857123,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 10947
+    },
+    {
+      "epoch": 0.5582763574418177,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 10948
+    },
+    {
+      "epoch": 0.5583273508979231,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 10949
+    },
+    {
+      "epoch": 0.5583783443540284,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5584293378101338,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 10951
+    },
+    {
+      "epoch": 0.5584803312662392,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 10952
+    },
+    {
+      "epoch": 0.5585313247223446,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 10953
+    },
+    {
+      "epoch": 0.55858231817845,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 10954
+    },
+    {
+      "epoch": 0.5586333116345554,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 10955
+    },
+    {
+      "epoch": 0.5586843050906608,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 10956
+    },
+    {
+      "epoch": 0.5587352985467662,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 10957
+    },
+    {
+      "epoch": 0.5587862920028716,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 10958
+    },
+    {
+      "epoch": 0.5588372854589769,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 10959
+    },
+    {
+      "epoch": 0.5588882789150823,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5589392723711878,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 10961
+    },
+    {
+      "epoch": 0.5589902658272932,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 10962
+    },
+    {
+      "epoch": 0.5590412592833985,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 10963
+    },
+    {
+      "epoch": 0.5590922527395039,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 10964
+    },
+    {
+      "epoch": 0.5591432461956093,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 10965
+    },
+    {
+      "epoch": 0.5591942396517147,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.774,
+      "step": 10966
+    },
+    {
+      "epoch": 0.55924523310782,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 10967
+    },
+    {
+      "epoch": 0.5592962265639255,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 10968
+    },
+    {
+      "epoch": 0.5593472200200309,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 10969
+    },
+    {
+      "epoch": 0.5593982134761363,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5594492069322416,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 10971
+    },
+    {
+      "epoch": 0.559500200388347,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 10972
+    },
+    {
+      "epoch": 0.5595511938444524,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10973
+    },
+    {
+      "epoch": 0.5596021873005578,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 10974
+    },
+    {
+      "epoch": 0.5596531807566633,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 10975
+    },
+    {
+      "epoch": 0.5597041742127686,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 10976
+    },
+    {
+      "epoch": 0.559755167668874,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 10977
+    },
+    {
+      "epoch": 0.5598061611249794,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 10978
+    },
+    {
+      "epoch": 0.5598571545810848,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 10979
+    },
+    {
+      "epoch": 0.5599081480371901,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5599591414932955,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 10981
+    },
+    {
+      "epoch": 0.560010134949401,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 10982
+    },
+    {
+      "epoch": 0.5600611284055064,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 10983
+    },
+    {
+      "epoch": 0.5601121218616117,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 10984
+    },
+    {
+      "epoch": 0.5601631153177171,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7622,
+      "step": 10985
+    },
+    {
+      "epoch": 0.5602141087738225,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 10986
+    },
+    {
+      "epoch": 0.5602651022299279,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 10987
+    },
+    {
+      "epoch": 0.5603160956860332,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7758,
+      "step": 10988
+    },
+    {
+      "epoch": 0.5603670891421387,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7689,
+      "step": 10989
+    },
+    {
+      "epoch": 0.5604180825982441,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5604690760543495,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 10991
+    },
+    {
+      "epoch": 0.5605200695104549,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 10992
+    },
+    {
+      "epoch": 0.5605710629665602,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 10993
+    },
+    {
+      "epoch": 0.5606220564226656,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 10994
+    },
+    {
+      "epoch": 0.560673049878771,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 10995
+    },
+    {
+      "epoch": 0.5607240433348765,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 10996
+    },
+    {
+      "epoch": 0.5607750367909818,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 10997
+    },
+    {
+      "epoch": 0.5608260302470872,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 10998
+    },
+    {
+      "epoch": 0.5608770237031926,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 10999
+    },
+    {
+      "epoch": 0.560928017159298,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5609790106154033,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 11001
+    },
+    {
+      "epoch": 0.5610300040715087,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 11002
+    },
+    {
+      "epoch": 0.5610809975276142,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 11003
+    },
+    {
+      "epoch": 0.5611319909837196,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 11004
+    },
+    {
+      "epoch": 0.561182984439825,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 11005
+    },
+    {
+      "epoch": 0.5612339778959303,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 11006
+    },
+    {
+      "epoch": 0.5612849713520357,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 11007
+    },
+    {
+      "epoch": 0.5613359648081411,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 11008
+    },
+    {
+      "epoch": 0.5613869582642464,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 11009
+    },
+    {
+      "epoch": 0.5614379517203519,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5614889451764573,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 11011
+    },
+    {
+      "epoch": 0.5615399386325627,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 11012
+    },
+    {
+      "epoch": 0.5615909320886681,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 11013
+    },
+    {
+      "epoch": 0.5616419255447734,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 11014
+    },
+    {
+      "epoch": 0.5616929190008788,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7893,
+      "step": 11015
+    },
+    {
+      "epoch": 0.5617439124569842,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 11016
+    },
+    {
+      "epoch": 0.5617949059130897,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 11017
+    },
+    {
+      "epoch": 0.561845899369195,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 11018
+    },
+    {
+      "epoch": 0.5618968928253004,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 11019
+    },
+    {
+      "epoch": 0.5619478862814058,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5619988797375112,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 11021
+    },
+    {
+      "epoch": 0.5620498731936165,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 11022
+    },
+    {
+      "epoch": 0.5621008666497219,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 11023
+    },
+    {
+      "epoch": 0.5621518601058274,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 11024
+    },
+    {
+      "epoch": 0.5622028535619328,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 11025
+    },
+    {
+      "epoch": 0.5622538470180382,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 11026
+    },
+    {
+      "epoch": 0.5623048404741435,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 11027
+    },
+    {
+      "epoch": 0.5623558339302489,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 11028
+    },
+    {
+      "epoch": 0.5624068273863543,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 11029
+    },
+    {
+      "epoch": 0.5624578208424597,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5625088142985651,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 11031
+    },
+    {
+      "epoch": 0.5625598077546705,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 11032
+    },
+    {
+      "epoch": 0.5626108012107759,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 11033
+    },
+    {
+      "epoch": 0.5626617946668813,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 11034
+    },
+    {
+      "epoch": 0.5627127881229866,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 11035
+    },
+    {
+      "epoch": 0.562763781579092,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 11036
+    },
+    {
+      "epoch": 0.5628147750351974,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 11037
+    },
+    {
+      "epoch": 0.5628657684913029,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 11038
+    },
+    {
+      "epoch": 0.5629167619474083,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7788,
+      "step": 11039
+    },
+    {
+      "epoch": 0.5629677554035136,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 11040
+    },
+    {
+      "epoch": 0.563018748859619,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 11041
+    },
+    {
+      "epoch": 0.5630697423157244,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 11042
+    },
+    {
+      "epoch": 0.5631207357718298,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 11043
+    },
+    {
+      "epoch": 0.5631717292279351,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 11044
+    },
+    {
+      "epoch": 0.5632227226840406,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7776,
+      "step": 11045
+    },
+    {
+      "epoch": 0.563273716140146,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 11046
+    },
+    {
+      "epoch": 0.5633247095962514,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7741,
+      "step": 11047
+    },
+    {
+      "epoch": 0.5633757030523567,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 11048
+    },
+    {
+      "epoch": 0.5634266965084621,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 11049
+    },
+    {
+      "epoch": 0.5634776899645675,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5635286834206729,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 11051
+    },
+    {
+      "epoch": 0.5635796768767783,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 11052
+    },
+    {
+      "epoch": 0.5636306703328837,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 11053
+    },
+    {
+      "epoch": 0.5636816637889891,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 11054
+    },
+    {
+      "epoch": 0.5637326572450945,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 11055
+    },
+    {
+      "epoch": 0.5637836507011998,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 11056
+    },
+    {
+      "epoch": 0.5638346441573052,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 11057
+    },
+    {
+      "epoch": 0.5638856376134106,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 11058
+    },
+    {
+      "epoch": 0.5639366310695161,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 11059
+    },
+    {
+      "epoch": 0.5639876245256215,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5640386179817268,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 11061
+    },
+    {
+      "epoch": 0.5640896114378322,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 11062
+    },
+    {
+      "epoch": 0.5641406048939376,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 11063
+    },
+    {
+      "epoch": 0.564191598350043,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 11064
+    },
+    {
+      "epoch": 0.5642425918061483,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 11065
+    },
+    {
+      "epoch": 0.5642935852622538,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 11066
+    },
+    {
+      "epoch": 0.5643445787183592,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 11067
+    },
+    {
+      "epoch": 0.5643955721744646,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 11068
+    },
+    {
+      "epoch": 0.5644465656305699,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 11069
+    },
+    {
+      "epoch": 0.5644975590866753,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5645485525427807,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 11071
+    },
+    {
+      "epoch": 0.5645995459988861,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 11072
+    },
+    {
+      "epoch": 0.5646505394549916,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 11073
+    },
+    {
+      "epoch": 0.5647015329110969,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 11074
+    },
+    {
+      "epoch": 0.5647525263672023,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11075
+    },
+    {
+      "epoch": 0.5648035198233077,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 11076
+    },
+    {
+      "epoch": 0.564854513279413,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 11077
+    },
+    {
+      "epoch": 0.5649055067355184,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 11078
+    },
+    {
+      "epoch": 0.5649565001916238,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 11079
+    },
+    {
+      "epoch": 0.5650074936477293,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5650584871038347,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11081
+    },
+    {
+      "epoch": 0.56510948055994,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 11082
+    },
+    {
+      "epoch": 0.5651604740160454,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 11083
+    },
+    {
+      "epoch": 0.5652114674721508,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 11084
+    },
+    {
+      "epoch": 0.5652624609282562,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 11085
+    },
+    {
+      "epoch": 0.5653134543843615,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11086
+    },
+    {
+      "epoch": 0.565364447840467,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 11087
+    },
+    {
+      "epoch": 0.5654154412965724,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 11088
+    },
+    {
+      "epoch": 0.5654664347526778,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 11089
+    },
+    {
+      "epoch": 0.5655174282087831,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5655684216648885,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 11091
+    },
+    {
+      "epoch": 0.5656194151209939,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7683,
+      "step": 11092
+    },
+    {
+      "epoch": 0.5656704085770993,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 11093
+    },
+    {
+      "epoch": 0.5657214020332048,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 11094
+    },
+    {
+      "epoch": 0.5657723954893101,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 11095
+    },
+    {
+      "epoch": 0.5658233889454155,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 11096
+    },
+    {
+      "epoch": 0.5658743824015209,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 11097
+    },
+    {
+      "epoch": 0.5659253758576263,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 11098
+    },
+    {
+      "epoch": 0.5659763693137316,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 11099
+    },
+    {
+      "epoch": 0.566027362769837,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5660783562259425,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 11101
+    },
+    {
+      "epoch": 0.5661293496820479,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 11102
+    },
+    {
+      "epoch": 0.5661803431381532,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7658,
+      "step": 11103
+    },
+    {
+      "epoch": 0.5662313365942586,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 11104
+    },
+    {
+      "epoch": 0.566282330050364,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 11105
+    },
+    {
+      "epoch": 0.5663333235064694,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 11106
+    },
+    {
+      "epoch": 0.5663843169625747,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7749,
+      "step": 11107
+    },
+    {
+      "epoch": 0.5664353104186802,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11108
+    },
+    {
+      "epoch": 0.5664863038747856,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 11109
+    },
+    {
+      "epoch": 0.566537297330891,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5665882907869964,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 11111
+    },
+    {
+      "epoch": 0.5666392842431017,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 11112
+    },
+    {
+      "epoch": 0.5666902776992071,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 11113
+    },
+    {
+      "epoch": 0.5667412711553125,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11114
+    },
+    {
+      "epoch": 0.566792264611418,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 11115
+    },
+    {
+      "epoch": 0.5668432580675233,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 11116
+    },
+    {
+      "epoch": 0.5668942515236287,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 11117
+    },
+    {
+      "epoch": 0.5669452449797341,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 11118
+    },
+    {
+      "epoch": 0.5669962384358395,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 11119
+    },
+    {
+      "epoch": 0.5670472318919448,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5670982253480502,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 11121
+    },
+    {
+      "epoch": 0.5671492188041557,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 11122
+    },
+    {
+      "epoch": 0.5672002122602611,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 11123
+    },
+    {
+      "epoch": 0.5672512057163664,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 11124
+    },
+    {
+      "epoch": 0.5673021991724718,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 11125
+    },
+    {
+      "epoch": 0.5673531926285772,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 11126
+    },
+    {
+      "epoch": 0.5674041860846826,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 11127
+    },
+    {
+      "epoch": 0.567455179540788,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 11128
+    },
+    {
+      "epoch": 0.5675061729968934,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 11129
+    },
+    {
+      "epoch": 0.5675571664529988,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5676081599091042,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 11131
+    },
+    {
+      "epoch": 0.5676591533652096,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 11132
+    },
+    {
+      "epoch": 0.5677101468213149,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11133
+    },
+    {
+      "epoch": 0.5677611402774203,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 11134
+    },
+    {
+      "epoch": 0.5678121337335257,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 11135
+    },
+    {
+      "epoch": 0.5678631271896312,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 11136
+    },
+    {
+      "epoch": 0.5679141206457365,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 11137
+    },
+    {
+      "epoch": 0.5679651141018419,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7752,
+      "step": 11138
+    },
+    {
+      "epoch": 0.5680161075579473,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 11139
+    },
+    {
+      "epoch": 0.5680671010140527,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 11140
+    },
+    {
+      "epoch": 0.568118094470158,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 11141
+    },
+    {
+      "epoch": 0.5681690879262634,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 11142
+    },
+    {
+      "epoch": 0.5682200813823689,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 11143
+    },
+    {
+      "epoch": 0.5682710748384743,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 11144
+    },
+    {
+      "epoch": 0.5683220682945797,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 11145
+    },
+    {
+      "epoch": 0.568373061750685,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 11146
+    },
+    {
+      "epoch": 0.5684240552067904,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8405,
+      "step": 11147
+    },
+    {
+      "epoch": 0.5684750486628958,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 11148
+    },
+    {
+      "epoch": 0.5685260421190012,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8361,
+      "step": 11149
+    },
+    {
+      "epoch": 0.5685770355751066,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 11150
+    },
+    {
+      "epoch": 0.568628029031212,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 11151
+    },
+    {
+      "epoch": 0.5686790224873174,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 11152
+    },
+    {
+      "epoch": 0.5687300159434228,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 11153
+    },
+    {
+      "epoch": 0.5687810093995281,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 11154
+    },
+    {
+      "epoch": 0.5688320028556335,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 11155
+    },
+    {
+      "epoch": 0.5688829963117389,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 11156
+    },
+    {
+      "epoch": 0.5689339897678444,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 11157
+    },
+    {
+      "epoch": 0.5689849832239497,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 11158
+    },
+    {
+      "epoch": 0.5690359766800551,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11159
+    },
+    {
+      "epoch": 0.5690869701361605,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5691379635922659,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 11161
+    },
+    {
+      "epoch": 0.5691889570483712,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 11162
+    },
+    {
+      "epoch": 0.5692399505044766,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 11163
+    },
+    {
+      "epoch": 0.5692909439605821,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 11164
+    },
+    {
+      "epoch": 0.5693419374166875,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 11165
+    },
+    {
+      "epoch": 0.5693929308727929,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 11166
+    },
+    {
+      "epoch": 0.5694439243288982,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 11167
+    },
+    {
+      "epoch": 0.5694949177850036,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 11168
+    },
+    {
+      "epoch": 0.569545911241109,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 11169
+    },
+    {
+      "epoch": 0.5695969046972144,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5696478981533198,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 11171
+    },
+    {
+      "epoch": 0.5696988916094252,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 11172
+    },
+    {
+      "epoch": 0.5697498850655306,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 11173
+    },
+    {
+      "epoch": 0.569800878521636,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11174
+    },
+    {
+      "epoch": 0.5698518719777413,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11175
+    },
+    {
+      "epoch": 0.5699028654338467,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 11176
+    },
+    {
+      "epoch": 0.5699538588899521,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 11177
+    },
+    {
+      "epoch": 0.5700048523460576,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 11178
+    },
+    {
+      "epoch": 0.570055845802163,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11179
+    },
+    {
+      "epoch": 0.5701068392582683,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7734,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5701578327143737,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 11181
+    },
+    {
+      "epoch": 0.5702088261704791,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 11182
+    },
+    {
+      "epoch": 0.5702598196265845,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 11183
+    },
+    {
+      "epoch": 0.5703108130826898,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7649,
+      "step": 11184
+    },
+    {
+      "epoch": 0.5703618065387953,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 11185
+    },
+    {
+      "epoch": 0.5704127999949007,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 11186
+    },
+    {
+      "epoch": 0.5704637934510061,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 11187
+    },
+    {
+      "epoch": 0.5705147869071114,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 11188
+    },
+    {
+      "epoch": 0.5705657803632168,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 11189
+    },
+    {
+      "epoch": 0.5706167738193222,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5706677672754276,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 11191
+    },
+    {
+      "epoch": 0.570718760731533,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7858,
+      "step": 11192
+    },
+    {
+      "epoch": 0.5707697541876384,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11193
+    },
+    {
+      "epoch": 0.5708207476437438,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 11194
+    },
+    {
+      "epoch": 0.5708717410998492,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 11195
+    },
+    {
+      "epoch": 0.5709227345559545,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 11196
+    },
+    {
+      "epoch": 0.5709737280120599,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 11197
+    },
+    {
+      "epoch": 0.5710247214681653,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 11198
+    },
+    {
+      "epoch": 0.5710757149242708,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 11199
+    },
+    {
+      "epoch": 0.5711267083803762,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5711777018364815,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 11201
+    },
+    {
+      "epoch": 0.5712286952925869,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 11202
+    },
+    {
+      "epoch": 0.5712796887486923,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 11203
+    },
+    {
+      "epoch": 0.5713306822047977,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 11204
+    },
+    {
+      "epoch": 0.571381675660903,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 11205
+    },
+    {
+      "epoch": 0.5714326691170085,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8542,
+      "step": 11206
+    },
+    {
+      "epoch": 0.5714836625731139,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 11207
+    },
+    {
+      "epoch": 0.5715346560292193,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 11208
+    },
+    {
+      "epoch": 0.5715856494853246,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 11209
+    },
+    {
+      "epoch": 0.57163664294143,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5716876363975354,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 11211
+    },
+    {
+      "epoch": 0.5717386298536408,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 11212
+    },
+    {
+      "epoch": 0.5717896233097463,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 11213
+    },
+    {
+      "epoch": 0.5718406167658516,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 11214
+    },
+    {
+      "epoch": 0.571891610221957,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 11215
+    },
+    {
+      "epoch": 0.5719426036780624,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 11216
+    },
+    {
+      "epoch": 0.5719935971341678,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 11217
+    },
+    {
+      "epoch": 0.5720445905902731,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11218
+    },
+    {
+      "epoch": 0.5720955840463785,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 11219
+    },
+    {
+      "epoch": 0.572146577502484,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5721975709585894,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 11221
+    },
+    {
+      "epoch": 0.5722485644146947,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 11222
+    },
+    {
+      "epoch": 0.5722995578708001,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 11223
+    },
+    {
+      "epoch": 0.5723505513269055,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 11224
+    },
+    {
+      "epoch": 0.5724015447830109,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 11225
+    },
+    {
+      "epoch": 0.5724525382391162,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7543,
+      "step": 11226
+    },
+    {
+      "epoch": 0.5725035316952217,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 11227
+    },
+    {
+      "epoch": 0.5725545251513271,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 11228
+    },
+    {
+      "epoch": 0.5726055186074325,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 11229
+    },
+    {
+      "epoch": 0.5726565120635378,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5727075055196432,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 11231
+    },
+    {
+      "epoch": 0.5727584989757486,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7865,
+      "step": 11232
+    },
+    {
+      "epoch": 0.572809492431854,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 11233
+    },
+    {
+      "epoch": 0.5728604858879595,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 11234
+    },
+    {
+      "epoch": 0.5729114793440648,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 11235
+    },
+    {
+      "epoch": 0.5729624728001702,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 11236
+    },
+    {
+      "epoch": 0.5730134662562756,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 11237
+    },
+    {
+      "epoch": 0.573064459712381,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11238
+    },
+    {
+      "epoch": 0.5731154531684863,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 11239
+    },
+    {
+      "epoch": 0.5731664466245917,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5732174400806972,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 11241
+    },
+    {
+      "epoch": 0.5732684335368026,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 11242
+    },
+    {
+      "epoch": 0.5733194269929079,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 11243
+    },
+    {
+      "epoch": 0.5733704204490133,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 11244
+    },
+    {
+      "epoch": 0.5734214139051187,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 11245
+    },
+    {
+      "epoch": 0.5734724073612241,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 11246
+    },
+    {
+      "epoch": 0.5735234008173294,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7785,
+      "step": 11247
+    },
+    {
+      "epoch": 0.5735743942734349,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 11248
+    },
+    {
+      "epoch": 0.5736253877295403,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 11249
+    },
+    {
+      "epoch": 0.5736763811856457,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 11250
+    },
+    {
+      "epoch": 0.573727374641751,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 11251
+    },
+    {
+      "epoch": 0.5737783680978564,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 11252
+    },
+    {
+      "epoch": 0.5738293615539618,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 11253
+    },
+    {
+      "epoch": 0.5738803550100672,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11254
+    },
+    {
+      "epoch": 0.5739313484661727,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 11255
+    },
+    {
+      "epoch": 0.573982341922278,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 11256
+    },
+    {
+      "epoch": 0.5740333353783834,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11257
+    },
+    {
+      "epoch": 0.5740843288344888,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 11258
+    },
+    {
+      "epoch": 0.5741353222905942,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 11259
+    },
+    {
+      "epoch": 0.5741863157466995,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5742373092028049,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 11261
+    },
+    {
+      "epoch": 0.5742883026589104,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 11262
+    },
+    {
+      "epoch": 0.5743392961150158,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 11263
+    },
+    {
+      "epoch": 0.5743902895711211,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11264
+    },
+    {
+      "epoch": 0.5744412830272265,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 11265
+    },
+    {
+      "epoch": 0.5744922764833319,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11266
+    },
+    {
+      "epoch": 0.5745432699394373,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 11267
+    },
+    {
+      "epoch": 0.5745942633955426,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 11268
+    },
+    {
+      "epoch": 0.5746452568516481,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 11269
+    },
+    {
+      "epoch": 0.5746962503077535,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5747472437638589,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 11271
+    },
+    {
+      "epoch": 0.5747982372199643,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 11272
+    },
+    {
+      "epoch": 0.5748492306760696,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 11273
+    },
+    {
+      "epoch": 0.574900224132175,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 11274
+    },
+    {
+      "epoch": 0.5749512175882804,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 11275
+    },
+    {
+      "epoch": 0.5750022110443859,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 11276
+    },
+    {
+      "epoch": 0.5750532045004912,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 11277
+    },
+    {
+      "epoch": 0.5751041979565966,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 11278
+    },
+    {
+      "epoch": 0.575155191412702,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7876,
+      "step": 11279
+    },
+    {
+      "epoch": 0.5752061848688074,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.776,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5752571783249127,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 11281
+    },
+    {
+      "epoch": 0.5753081717810181,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 11282
+    },
+    {
+      "epoch": 0.5753591652371236,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 11283
+    },
+    {
+      "epoch": 0.575410158693229,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 11284
+    },
+    {
+      "epoch": 0.5754611521493344,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 11285
+    },
+    {
+      "epoch": 0.5755121456054397,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 11286
+    },
+    {
+      "epoch": 0.5755631390615451,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 11287
+    },
+    {
+      "epoch": 0.5756141325176505,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 11288
+    },
+    {
+      "epoch": 0.5756651259737559,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 11289
+    },
+    {
+      "epoch": 0.5757161194298613,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5757671128859667,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 11291
+    },
+    {
+      "epoch": 0.5758181063420721,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 11292
+    },
+    {
+      "epoch": 0.5758690997981775,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 11293
+    },
+    {
+      "epoch": 0.5759200932542828,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 11294
+    },
+    {
+      "epoch": 0.5759710867103882,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7858,
+      "step": 11295
+    },
+    {
+      "epoch": 0.5760220801664936,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7676,
+      "step": 11296
+    },
+    {
+      "epoch": 0.5760730736225991,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 11297
+    },
+    {
+      "epoch": 0.5761240670787044,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 11298
+    },
+    {
+      "epoch": 0.5761750605348098,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 11299
+    },
+    {
+      "epoch": 0.5762260539909152,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5762770474470206,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 11301
+    },
+    {
+      "epoch": 0.576328040903126,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 11302
+    },
+    {
+      "epoch": 0.5763790343592313,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 11303
+    },
+    {
+      "epoch": 0.5764300278153368,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 11304
+    },
+    {
+      "epoch": 0.5764810212714422,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 11305
+    },
+    {
+      "epoch": 0.5765320147275476,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.773,
+      "step": 11306
+    },
+    {
+      "epoch": 0.5765830081836529,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 11307
+    },
+    {
+      "epoch": 0.5766340016397583,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 11308
+    },
+    {
+      "epoch": 0.5766849950958637,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7802,
+      "step": 11309
+    },
+    {
+      "epoch": 0.5767359885519691,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5767869820080745,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 11311
+    },
+    {
+      "epoch": 0.5768379754641799,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7607,
+      "step": 11312
+    },
+    {
+      "epoch": 0.5768889689202853,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 11313
+    },
+    {
+      "epoch": 0.5769399623763907,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 11314
+    },
+    {
+      "epoch": 0.576990955832496,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 11315
+    },
+    {
+      "epoch": 0.5770419492886014,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 11316
+    },
+    {
+      "epoch": 0.5770929427447068,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 11317
+    },
+    {
+      "epoch": 0.5771439362008123,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 11318
+    },
+    {
+      "epoch": 0.5771949296569177,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 11319
+    },
+    {
+      "epoch": 0.577245923113023,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5772969165691284,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 11321
+    },
+    {
+      "epoch": 0.5773479100252338,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 11322
+    },
+    {
+      "epoch": 0.5773989034813392,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 11323
+    },
+    {
+      "epoch": 0.5774498969374445,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 11324
+    },
+    {
+      "epoch": 0.57750089039355,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11325
+    },
+    {
+      "epoch": 0.5775518838496554,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 11326
+    },
+    {
+      "epoch": 0.5776028773057608,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 11327
+    },
+    {
+      "epoch": 0.5776538707618661,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 11328
+    },
+    {
+      "epoch": 0.5777048642179715,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 11329
+    },
+    {
+      "epoch": 0.5777558576740769,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5778068511301823,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 11331
+    },
+    {
+      "epoch": 0.5778578445862877,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 11332
+    },
+    {
+      "epoch": 0.5779088380423931,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 11333
+    },
+    {
+      "epoch": 0.5779598314984985,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 11334
+    },
+    {
+      "epoch": 0.5780108249546039,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 11335
+    },
+    {
+      "epoch": 0.5780618184107092,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 11336
+    },
+    {
+      "epoch": 0.5781128118668146,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 11337
+    },
+    {
+      "epoch": 0.57816380532292,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 11338
+    },
+    {
+      "epoch": 0.5782147987790255,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 11339
+    },
+    {
+      "epoch": 0.5782657922351309,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5783167856912362,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 11341
+    },
+    {
+      "epoch": 0.5783677791473416,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7785,
+      "step": 11342
+    },
+    {
+      "epoch": 0.578418772603447,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 11343
+    },
+    {
+      "epoch": 0.5784697660595524,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 11344
+    },
+    {
+      "epoch": 0.5785207595156577,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 11345
+    },
+    {
+      "epoch": 0.5785717529717632,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 11346
+    },
+    {
+      "epoch": 0.5786227464278686,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 11347
+    },
+    {
+      "epoch": 0.578673739883974,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 11348
+    },
+    {
+      "epoch": 0.5787247333400793,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 11349
+    },
+    {
+      "epoch": 0.5787757267961847,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5788267202522901,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 11351
+    },
+    {
+      "epoch": 0.5788777137083955,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 11352
+    },
+    {
+      "epoch": 0.578928707164501,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 11353
+    },
+    {
+      "epoch": 0.5789797006206063,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 11354
+    },
+    {
+      "epoch": 0.5790306940767117,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11355
+    },
+    {
+      "epoch": 0.5790816875328171,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 11356
+    },
+    {
+      "epoch": 0.5791326809889225,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 11357
+    },
+    {
+      "epoch": 0.5791836744450278,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 11358
+    },
+    {
+      "epoch": 0.5792346679011332,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 11359
+    },
+    {
+      "epoch": 0.5792856613572387,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5793366548133441,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 11361
+    },
+    {
+      "epoch": 0.5793876482694494,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 11362
+    },
+    {
+      "epoch": 0.5794386417255548,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7736,
+      "step": 11363
+    },
+    {
+      "epoch": 0.5794896351816602,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 11364
+    },
+    {
+      "epoch": 0.5795406286377656,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 11365
+    },
+    {
+      "epoch": 0.5795916220938709,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 11366
+    },
+    {
+      "epoch": 0.5796426155499764,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 11367
+    },
+    {
+      "epoch": 0.5796936090060818,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 11368
+    },
+    {
+      "epoch": 0.5797446024621872,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 11369
+    },
+    {
+      "epoch": 0.5797955959182926,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5798465893743979,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 11371
+    },
+    {
+      "epoch": 0.5798975828305033,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 11372
+    },
+    {
+      "epoch": 0.5799485762866087,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 11373
+    },
+    {
+      "epoch": 0.5799995697427142,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7717,
+      "step": 11374
+    },
+    {
+      "epoch": 0.5800505631988195,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 11375
+    },
+    {
+      "epoch": 0.5801015566549249,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 11376
+    },
+    {
+      "epoch": 0.5801525501110303,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 11377
+    },
+    {
+      "epoch": 0.5802035435671357,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 11378
+    },
+    {
+      "epoch": 0.580254537023241,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 11379
+    },
+    {
+      "epoch": 0.5803055304793464,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5803565239354519,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 11381
+    },
+    {
+      "epoch": 0.5804075173915573,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 11382
+    },
+    {
+      "epoch": 0.5804585108476626,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11383
+    },
+    {
+      "epoch": 0.580509504303768,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 11384
+    },
+    {
+      "epoch": 0.5805604977598734,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 11385
+    },
+    {
+      "epoch": 0.5806114912159788,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 11386
+    },
+    {
+      "epoch": 0.5806624846720841,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 11387
+    },
+    {
+      "epoch": 0.5807134781281896,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 11388
+    },
+    {
+      "epoch": 0.580764471584295,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 11389
+    },
+    {
+      "epoch": 0.5808154650404004,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5808664584965058,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 11391
+    },
+    {
+      "epoch": 0.5809174519526111,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 11392
+    },
+    {
+      "epoch": 0.5809684454087165,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 11393
+    },
+    {
+      "epoch": 0.5810194388648219,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 11394
+    },
+    {
+      "epoch": 0.5810704323209274,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 11395
+    },
+    {
+      "epoch": 0.5811214257770327,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 11396
+    },
+    {
+      "epoch": 0.5811724192331381,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 11397
+    },
+    {
+      "epoch": 0.5812234126892435,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11398
+    },
+    {
+      "epoch": 0.5812744061453489,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 11399
+    },
+    {
+      "epoch": 0.5813253996014542,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5813763930575596,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 11401
+    },
+    {
+      "epoch": 0.5814273865136651,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 11402
+    },
+    {
+      "epoch": 0.5814783799697705,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 11403
+    },
+    {
+      "epoch": 0.5815293734258759,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 11404
+    },
+    {
+      "epoch": 0.5815803668819812,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 11405
+    },
+    {
+      "epoch": 0.5816313603380866,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 11406
+    },
+    {
+      "epoch": 0.581682353794192,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 11407
+    },
+    {
+      "epoch": 0.5817333472502974,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.777,
+      "step": 11408
+    },
+    {
+      "epoch": 0.5817843407064028,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 11409
+    },
+    {
+      "epoch": 0.5818353341625082,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5818863276186136,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 11411
+    },
+    {
+      "epoch": 0.581937321074719,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 11412
+    },
+    {
+      "epoch": 0.5819883145308243,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 11413
+    },
+    {
+      "epoch": 0.5820393079869297,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7866,
+      "step": 11414
+    },
+    {
+      "epoch": 0.5820903014430351,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 11415
+    },
+    {
+      "epoch": 0.5821412948991406,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 11416
+    },
+    {
+      "epoch": 0.582192288355246,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 11417
+    },
+    {
+      "epoch": 0.5822432818113513,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 11418
+    },
+    {
+      "epoch": 0.5822942752674567,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 11419
+    },
+    {
+      "epoch": 0.5823452687235621,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5823962621796674,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 11421
+    },
+    {
+      "epoch": 0.5824472556357728,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 11422
+    },
+    {
+      "epoch": 0.5824982490918783,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 11423
+    },
+    {
+      "epoch": 0.5825492425479837,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 11424
+    },
+    {
+      "epoch": 0.5826002360040891,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 11425
+    },
+    {
+      "epoch": 0.5826512294601944,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 11426
+    },
+    {
+      "epoch": 0.5827022229162998,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 11427
+    },
+    {
+      "epoch": 0.5827532163724052,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7808,
+      "step": 11428
+    },
+    {
+      "epoch": 0.5828042098285106,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 11429
+    },
+    {
+      "epoch": 0.582855203284616,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5829061967407214,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 11431
+    },
+    {
+      "epoch": 0.5829571901968268,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 11432
+    },
+    {
+      "epoch": 0.5830081836529322,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 11433
+    },
+    {
+      "epoch": 0.5830591771090375,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 11434
+    },
+    {
+      "epoch": 0.5831101705651429,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 11435
+    },
+    {
+      "epoch": 0.5831611640212483,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11436
+    },
+    {
+      "epoch": 0.5832121574773538,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 11437
+    },
+    {
+      "epoch": 0.5832631509334592,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 11438
+    },
+    {
+      "epoch": 0.5833141443895645,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 11439
+    },
+    {
+      "epoch": 0.5833651378456699,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5834161313017753,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 11441
+    },
+    {
+      "epoch": 0.5834671247578807,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7823,
+      "step": 11442
+    },
+    {
+      "epoch": 0.583518118213986,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11443
+    },
+    {
+      "epoch": 0.5835691116700915,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 11444
+    },
+    {
+      "epoch": 0.5836201051261969,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 11445
+    },
+    {
+      "epoch": 0.5836710985823023,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 11446
+    },
+    {
+      "epoch": 0.5837220920384076,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7781,
+      "step": 11447
+    },
+    {
+      "epoch": 0.583773085494513,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7721,
+      "step": 11448
+    },
+    {
+      "epoch": 0.5838240789506184,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 11449
+    },
+    {
+      "epoch": 0.5838750724067238,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5839260658628292,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 11451
+    },
+    {
+      "epoch": 0.5839770593189346,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 11452
+    },
+    {
+      "epoch": 0.58402805277504,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 11453
+    },
+    {
+      "epoch": 0.5840790462311454,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 11454
+    },
+    {
+      "epoch": 0.5841300396872507,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7805,
+      "step": 11455
+    },
+    {
+      "epoch": 0.5841810331433561,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 11456
+    },
+    {
+      "epoch": 0.5842320265994615,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 11457
+    },
+    {
+      "epoch": 0.584283020055567,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 11458
+    },
+    {
+      "epoch": 0.5843340135116724,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 11459
+    },
+    {
+      "epoch": 0.5843850069677777,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5844360004238831,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 11461
+    },
+    {
+      "epoch": 0.5844869938799885,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 11462
+    },
+    {
+      "epoch": 0.5845379873360939,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 11463
+    },
+    {
+      "epoch": 0.5845889807921992,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7697,
+      "step": 11464
+    },
+    {
+      "epoch": 0.5846399742483047,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 11465
+    },
+    {
+      "epoch": 0.5846909677044101,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7815,
+      "step": 11466
+    },
+    {
+      "epoch": 0.5847419611605155,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 11467
+    },
+    {
+      "epoch": 0.5847929546166208,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 11468
+    },
+    {
+      "epoch": 0.5848439480727262,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 11469
+    },
+    {
+      "epoch": 0.5848949415288316,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7783,
+      "step": 11470
+    },
+    {
+      "epoch": 0.584945934984937,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 11471
+    },
+    {
+      "epoch": 0.5849969284410425,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 11472
+    },
+    {
+      "epoch": 0.5850479218971478,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 11473
+    },
+    {
+      "epoch": 0.5850989153532532,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 11474
+    },
+    {
+      "epoch": 0.5851499088093586,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 11475
+    },
+    {
+      "epoch": 0.585200902265464,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 11476
+    },
+    {
+      "epoch": 0.5852518957215693,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 11477
+    },
+    {
+      "epoch": 0.5853028891776747,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 11478
+    },
+    {
+      "epoch": 0.5853538826337802,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 11479
+    },
+    {
+      "epoch": 0.5854048760898856,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5854558695459909,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7741,
+      "step": 11481
+    },
+    {
+      "epoch": 0.5855068630020963,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 11482
+    },
+    {
+      "epoch": 0.5855578564582017,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 11483
+    },
+    {
+      "epoch": 0.5856088499143071,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11484
+    },
+    {
+      "epoch": 0.5856598433704124,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 11485
+    },
+    {
+      "epoch": 0.5857108368265179,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 11486
+    },
+    {
+      "epoch": 0.5857618302826233,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 11487
+    },
+    {
+      "epoch": 0.5858128237387287,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7606,
+      "step": 11488
+    },
+    {
+      "epoch": 0.585863817194834,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 11489
+    },
+    {
+      "epoch": 0.5859148106509394,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7823,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5859658041070448,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 11491
+    },
+    {
+      "epoch": 0.5860167975631502,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 11492
+    },
+    {
+      "epoch": 0.5860677910192557,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 11493
+    },
+    {
+      "epoch": 0.586118784475361,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 11494
+    },
+    {
+      "epoch": 0.5861697779314664,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 11495
+    },
+    {
+      "epoch": 0.5862207713875718,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 11496
+    },
+    {
+      "epoch": 0.5862717648436772,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 11497
+    },
+    {
+      "epoch": 0.5863227582997825,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 11498
+    },
+    {
+      "epoch": 0.5863737517558879,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 11499
+    },
+    {
+      "epoch": 0.5864247452119934,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5864757386680988,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 11501
+    },
+    {
+      "epoch": 0.5865267321242041,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 11502
+    },
+    {
+      "epoch": 0.5865777255803095,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 11503
+    },
+    {
+      "epoch": 0.5866287190364149,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11504
+    },
+    {
+      "epoch": 0.5866797124925203,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 11505
+    },
+    {
+      "epoch": 0.5867307059486256,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 11506
+    },
+    {
+      "epoch": 0.5867816994047311,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 11507
+    },
+    {
+      "epoch": 0.5868326928608365,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 11508
+    },
+    {
+      "epoch": 0.5868836863169419,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 11509
+    },
+    {
+      "epoch": 0.5869346797730473,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7705,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5869856732291526,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 11511
+    },
+    {
+      "epoch": 0.587036666685258,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 11512
+    },
+    {
+      "epoch": 0.5870876601413634,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 11513
+    },
+    {
+      "epoch": 0.5871386535974689,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 11514
+    },
+    {
+      "epoch": 0.5871896470535742,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 11515
+    },
+    {
+      "epoch": 0.5872406405096796,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 11516
+    },
+    {
+      "epoch": 0.587291633965785,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 11517
+    },
+    {
+      "epoch": 0.5873426274218904,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 11518
+    },
+    {
+      "epoch": 0.5873936208779957,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 11519
+    },
+    {
+      "epoch": 0.5874446143341011,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5874956077902066,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 11521
+    },
+    {
+      "epoch": 0.587546601246312,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 11522
+    },
+    {
+      "epoch": 0.5875975947024173,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 11523
+    },
+    {
+      "epoch": 0.5876485881585227,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11524
+    },
+    {
+      "epoch": 0.5876995816146281,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 11525
+    },
+    {
+      "epoch": 0.5877505750707335,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11526
+    },
+    {
+      "epoch": 0.5878015685268388,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 11527
+    },
+    {
+      "epoch": 0.5878525619829443,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 11528
+    },
+    {
+      "epoch": 0.5879035554390497,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8481,
+      "step": 11529
+    },
+    {
+      "epoch": 0.5879545488951551,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5880055423512605,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 11531
+    },
+    {
+      "epoch": 0.5880565358073658,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11532
+    },
+    {
+      "epoch": 0.5881075292634712,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 11533
+    },
+    {
+      "epoch": 0.5881585227195766,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 11534
+    },
+    {
+      "epoch": 0.5882095161756821,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 11535
+    },
+    {
+      "epoch": 0.5882605096317874,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 11536
+    },
+    {
+      "epoch": 0.5883115030878928,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 11537
+    },
+    {
+      "epoch": 0.5883624965439982,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 11538
+    },
+    {
+      "epoch": 0.5884134900001036,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 11539
+    },
+    {
+      "epoch": 0.5884644834562089,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5885154769123143,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 11541
+    },
+    {
+      "epoch": 0.5885664703684198,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 11542
+    },
+    {
+      "epoch": 0.5886174638245252,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 11543
+    },
+    {
+      "epoch": 0.5886684572806306,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 11544
+    },
+    {
+      "epoch": 0.5887194507367359,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 11545
+    },
+    {
+      "epoch": 0.5887704441928413,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11546
+    },
+    {
+      "epoch": 0.5888214376489467,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 11547
+    },
+    {
+      "epoch": 0.588872431105052,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11548
+    },
+    {
+      "epoch": 0.5889234245611575,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 11549
+    },
+    {
+      "epoch": 0.5889744180172629,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5890254114733683,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 11551
+    },
+    {
+      "epoch": 0.5890764049294737,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 11552
+    },
+    {
+      "epoch": 0.589127398385579,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 11553
+    },
+    {
+      "epoch": 0.5891783918416844,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 11554
+    },
+    {
+      "epoch": 0.5892293852977898,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 11555
+    },
+    {
+      "epoch": 0.5892803787538953,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 11556
+    },
+    {
+      "epoch": 0.5893313722100006,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 11557
+    },
+    {
+      "epoch": 0.589382365666106,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 11558
+    },
+    {
+      "epoch": 0.5894333591222114,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 11559
+    },
+    {
+      "epoch": 0.5894843525783168,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5895353460344221,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 11561
+    },
+    {
+      "epoch": 0.5895863394905275,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 11562
+    },
+    {
+      "epoch": 0.589637332946633,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 11563
+    },
+    {
+      "epoch": 0.5896883264027384,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 11564
+    },
+    {
+      "epoch": 0.5897393198588438,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 11565
+    },
+    {
+      "epoch": 0.5897903133149491,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 11566
+    },
+    {
+      "epoch": 0.5898413067710545,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11567
+    },
+    {
+      "epoch": 0.5898923002271599,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 11568
+    },
+    {
+      "epoch": 0.5899432936832653,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 11569
+    },
+    {
+      "epoch": 0.5899942871393707,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5900452805954761,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 11571
+    },
+    {
+      "epoch": 0.5900962740515815,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11572
+    },
+    {
+      "epoch": 0.5901472675076869,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 11573
+    },
+    {
+      "epoch": 0.5901982609637922,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 11574
+    },
+    {
+      "epoch": 0.5902492544198976,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 11575
+    },
+    {
+      "epoch": 0.590300247876003,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 11576
+    },
+    {
+      "epoch": 0.5903512413321085,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7805,
+      "step": 11577
+    },
+    {
+      "epoch": 0.5904022347882139,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 11578
+    },
+    {
+      "epoch": 0.5904532282443192,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 11579
+    },
+    {
+      "epoch": 0.5905042217004246,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 11580
+    },
+    {
+      "epoch": 0.59055521515653,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.77,
+      "step": 11581
+    },
+    {
+      "epoch": 0.5906062086126354,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 11582
+    },
+    {
+      "epoch": 0.5906572020687407,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 11583
+    },
+    {
+      "epoch": 0.5907081955248462,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 11584
+    },
+    {
+      "epoch": 0.5907591889809516,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 11585
+    },
+    {
+      "epoch": 0.590810182437057,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 11586
+    },
+    {
+      "epoch": 0.5908611758931623,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 11587
+    },
+    {
+      "epoch": 0.5909121693492677,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 11588
+    },
+    {
+      "epoch": 0.5909631628053731,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 11589
+    },
+    {
+      "epoch": 0.5910141562614785,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 11590
+    },
+    {
+      "epoch": 0.591065149717584,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7835,
+      "step": 11591
+    },
+    {
+      "epoch": 0.5911161431736893,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 11592
+    },
+    {
+      "epoch": 0.5911671366297947,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 11593
+    },
+    {
+      "epoch": 0.5912181300859001,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7763,
+      "step": 11594
+    },
+    {
+      "epoch": 0.5912691235420054,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 11595
+    },
+    {
+      "epoch": 0.5913201169981108,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 11596
+    },
+    {
+      "epoch": 0.5913711104542162,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11597
+    },
+    {
+      "epoch": 0.5914221039103217,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 11598
+    },
+    {
+      "epoch": 0.5914730973664271,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 11599
+    },
+    {
+      "epoch": 0.5915240908225324,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5915750842786378,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 11601
+    },
+    {
+      "epoch": 0.5916260777347432,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 11602
+    },
+    {
+      "epoch": 0.5916770711908486,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 11603
+    },
+    {
+      "epoch": 0.5917280646469539,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 11604
+    },
+    {
+      "epoch": 0.5917790581030594,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 11605
+    },
+    {
+      "epoch": 0.5918300515591648,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 11606
+    },
+    {
+      "epoch": 0.5918810450152702,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11607
+    },
+    {
+      "epoch": 0.5919320384713755,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7797,
+      "step": 11608
+    },
+    {
+      "epoch": 0.5919830319274809,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 11609
+    },
+    {
+      "epoch": 0.5920340253835863,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7797,
+      "step": 11610
+    },
+    {
+      "epoch": 0.5920850188396917,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 11611
+    },
+    {
+      "epoch": 0.5921360122957972,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 11612
+    },
+    {
+      "epoch": 0.5921870057519025,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 11613
+    },
+    {
+      "epoch": 0.5922379992080079,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7638,
+      "step": 11614
+    },
+    {
+      "epoch": 0.5922889926641133,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 11615
+    },
+    {
+      "epoch": 0.5923399861202187,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 11616
+    },
+    {
+      "epoch": 0.592390979576324,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 11617
+    },
+    {
+      "epoch": 0.5924419730324294,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 11618
+    },
+    {
+      "epoch": 0.5924929664885349,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 11619
+    },
+    {
+      "epoch": 0.5925439599446403,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5925949534007456,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 11621
+    },
+    {
+      "epoch": 0.592645946856851,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 11622
+    },
+    {
+      "epoch": 0.5926969403129564,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 11623
+    },
+    {
+      "epoch": 0.5927479337690618,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 11624
+    },
+    {
+      "epoch": 0.5927989272251671,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 11625
+    },
+    {
+      "epoch": 0.5928499206812726,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 11626
+    },
+    {
+      "epoch": 0.592900914137378,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7672,
+      "step": 11627
+    },
+    {
+      "epoch": 0.5929519075934834,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 11628
+    },
+    {
+      "epoch": 0.5930029010495887,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 11629
+    },
+    {
+      "epoch": 0.5930538945056941,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5931048879617995,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 11631
+    },
+    {
+      "epoch": 0.5931558814179049,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 11632
+    },
+    {
+      "epoch": 0.5932068748740104,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 11633
+    },
+    {
+      "epoch": 0.5932578683301157,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11634
+    },
+    {
+      "epoch": 0.5933088617862211,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 11635
+    },
+    {
+      "epoch": 0.5933598552423265,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 11636
+    },
+    {
+      "epoch": 0.5934108486984319,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 11637
+    },
+    {
+      "epoch": 0.5934618421545372,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 11638
+    },
+    {
+      "epoch": 0.5935128356106426,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 11639
+    },
+    {
+      "epoch": 0.5935638290667481,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5936148225228535,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 11641
+    },
+    {
+      "epoch": 0.5936658159789588,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 11642
+    },
+    {
+      "epoch": 0.5937168094350642,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 11643
+    },
+    {
+      "epoch": 0.5937678028911696,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 11644
+    },
+    {
+      "epoch": 0.593818796347275,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 11645
+    },
+    {
+      "epoch": 0.5938697898033803,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 11646
+    },
+    {
+      "epoch": 0.5939207832594858,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 11647
+    },
+    {
+      "epoch": 0.5939717767155912,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 11648
+    },
+    {
+      "epoch": 0.5940227701716966,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 11649
+    },
+    {
+      "epoch": 0.594073763627802,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5941247570839073,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 11651
+    },
+    {
+      "epoch": 0.5941757505400127,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 11652
+    },
+    {
+      "epoch": 0.5942267439961181,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11653
+    },
+    {
+      "epoch": 0.5942777374522236,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 11654
+    },
+    {
+      "epoch": 0.5943287309083289,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 11655
+    },
+    {
+      "epoch": 0.5943797243644343,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 11656
+    },
+    {
+      "epoch": 0.5944307178205397,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11657
+    },
+    {
+      "epoch": 0.5944817112766451,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 11658
+    },
+    {
+      "epoch": 0.5945327047327504,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 11659
+    },
+    {
+      "epoch": 0.5945836981888558,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5946346916449613,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.831,
+      "step": 11661
+    },
+    {
+      "epoch": 0.5946856851010667,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7709,
+      "step": 11662
+    },
+    {
+      "epoch": 0.594736678557172,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 11663
+    },
+    {
+      "epoch": 0.5947876720132774,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 11664
+    },
+    {
+      "epoch": 0.5948386654693828,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 11665
+    },
+    {
+      "epoch": 0.5948896589254882,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 11666
+    },
+    {
+      "epoch": 0.5949406523815935,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 11667
+    },
+    {
+      "epoch": 0.594991645837699,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 11668
+    },
+    {
+      "epoch": 0.5950426392938044,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7829,
+      "step": 11669
+    },
+    {
+      "epoch": 0.5950936327499098,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5951446262060152,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 11671
+    },
+    {
+      "epoch": 0.5951956196621205,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 11672
+    },
+    {
+      "epoch": 0.5952466131182259,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7756,
+      "step": 11673
+    },
+    {
+      "epoch": 0.5952976065743313,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 11674
+    },
+    {
+      "epoch": 0.5953486000304368,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 11675
+    },
+    {
+      "epoch": 0.5953995934865421,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 11676
+    },
+    {
+      "epoch": 0.5954505869426475,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 11677
+    },
+    {
+      "epoch": 0.5955015803987529,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 11678
+    },
+    {
+      "epoch": 0.5955525738548583,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 11679
+    },
+    {
+      "epoch": 0.5956035673109636,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8366,
+      "step": 11680
+    },
+    {
+      "epoch": 0.595654560767069,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 11681
+    },
+    {
+      "epoch": 0.5957055542231745,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 11682
+    },
+    {
+      "epoch": 0.5957565476792799,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 11683
+    },
+    {
+      "epoch": 0.5958075411353853,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 11684
+    },
+    {
+      "epoch": 0.5958585345914906,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 11685
+    },
+    {
+      "epoch": 0.595909528047596,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 11686
+    },
+    {
+      "epoch": 0.5959605215037014,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 11687
+    },
+    {
+      "epoch": 0.5960115149598068,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.775,
+      "step": 11688
+    },
+    {
+      "epoch": 0.5960625084159122,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 11689
+    },
+    {
+      "epoch": 0.5961135018720176,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 11690
+    },
+    {
+      "epoch": 0.596164495328123,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 11691
+    },
+    {
+      "epoch": 0.5962154887842284,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7706,
+      "step": 11692
+    },
+    {
+      "epoch": 0.5962664822403337,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 11693
+    },
+    {
+      "epoch": 0.5963174756964391,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 11694
+    },
+    {
+      "epoch": 0.5963684691525445,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 11695
+    },
+    {
+      "epoch": 0.59641946260865,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 11696
+    },
+    {
+      "epoch": 0.5964704560647553,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 11697
+    },
+    {
+      "epoch": 0.5965214495208607,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 11698
+    },
+    {
+      "epoch": 0.5965724429769661,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 11699
+    },
+    {
+      "epoch": 0.5966234364330715,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5966744298891768,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 11701
+    },
+    {
+      "epoch": 0.5967254233452822,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 11702
+    },
+    {
+      "epoch": 0.5967764168013877,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 11703
+    },
+    {
+      "epoch": 0.5968274102574931,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.772,
+      "step": 11704
+    },
+    {
+      "epoch": 0.5968784037135985,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.768,
+      "step": 11705
+    },
+    {
+      "epoch": 0.5969293971697038,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7858,
+      "step": 11706
+    },
+    {
+      "epoch": 0.5969803906258092,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11707
+    },
+    {
+      "epoch": 0.5970313840819146,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.777,
+      "step": 11708
+    },
+    {
+      "epoch": 0.59708237753802,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11709
+    },
+    {
+      "epoch": 0.5971333709941254,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5971843644502308,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 11711
+    },
+    {
+      "epoch": 0.5972353579063362,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11712
+    },
+    {
+      "epoch": 0.5972863513624416,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 11713
+    },
+    {
+      "epoch": 0.597337344818547,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 11714
+    },
+    {
+      "epoch": 0.5973883382746523,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 11715
+    },
+    {
+      "epoch": 0.5974393317307577,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 11716
+    },
+    {
+      "epoch": 0.5974903251868632,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 11717
+    },
+    {
+      "epoch": 0.5975413186429686,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 11718
+    },
+    {
+      "epoch": 0.5975923120990739,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 11719
+    },
+    {
+      "epoch": 0.5976433055551793,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5976942990112847,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 11721
+    },
+    {
+      "epoch": 0.59774529246739,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 11722
+    },
+    {
+      "epoch": 0.5977962859234954,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 11723
+    },
+    {
+      "epoch": 0.5978472793796009,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 11724
+    },
+    {
+      "epoch": 0.5978982728357063,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 11725
+    },
+    {
+      "epoch": 0.5979492662918117,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7774,
+      "step": 11726
+    },
+    {
+      "epoch": 0.598000259747917,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 11727
+    },
+    {
+      "epoch": 0.5980512532040224,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 11728
+    },
+    {
+      "epoch": 0.5981022466601278,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 11729
+    },
+    {
+      "epoch": 0.5981532401162332,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7721,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5982042335723387,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11731
+    },
+    {
+      "epoch": 0.598255227028444,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.782,
+      "step": 11732
+    },
+    {
+      "epoch": 0.5983062204845494,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 11733
+    },
+    {
+      "epoch": 0.5983572139406548,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 11734
+    },
+    {
+      "epoch": 0.5984082073967602,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 11735
+    },
+    {
+      "epoch": 0.5984592008528655,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 11736
+    },
+    {
+      "epoch": 0.5985101943089709,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 11737
+    },
+    {
+      "epoch": 0.5985611877650764,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 11738
+    },
+    {
+      "epoch": 0.5986121812211818,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 11739
+    },
+    {
+      "epoch": 0.5986631746772871,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5987141681333925,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 11741
+    },
+    {
+      "epoch": 0.5987651615894979,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 11742
+    },
+    {
+      "epoch": 0.5988161550456033,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.779,
+      "step": 11743
+    },
+    {
+      "epoch": 0.5988671485017086,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 11744
+    },
+    {
+      "epoch": 0.5989181419578141,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 11745
+    },
+    {
+      "epoch": 0.5989691354139195,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 11746
+    },
+    {
+      "epoch": 0.5990201288700249,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 11747
+    },
+    {
+      "epoch": 0.5990711223261302,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 11748
+    },
+    {
+      "epoch": 0.5991221157822356,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 11749
+    },
+    {
+      "epoch": 0.599173109238341,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5992241026944464,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7711,
+      "step": 11751
+    },
+    {
+      "epoch": 0.5992750961505519,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 11752
+    },
+    {
+      "epoch": 0.5993260896066572,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 11753
+    },
+    {
+      "epoch": 0.5993770830627626,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 11754
+    },
+    {
+      "epoch": 0.599428076518868,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 11755
+    },
+    {
+      "epoch": 0.5994790699749734,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 11756
+    },
+    {
+      "epoch": 0.5995300634310787,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 11757
+    },
+    {
+      "epoch": 0.5995810568871841,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 11758
+    },
+    {
+      "epoch": 0.5996320503432896,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 11759
+    },
+    {
+      "epoch": 0.599683043799395,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5997340372555003,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11761
+    },
+    {
+      "epoch": 0.5997850307116057,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 11762
+    },
+    {
+      "epoch": 0.5998360241677111,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 11763
+    },
+    {
+      "epoch": 0.5998870176238165,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11764
+    },
+    {
+      "epoch": 0.5999380110799218,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 11765
+    },
+    {
+      "epoch": 0.5999890045360273,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 11766
+    },
+    {
+      "epoch": 0.6000399979921327,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 11767
+    },
+    {
+      "epoch": 0.6000909914482381,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 11768
+    },
+    {
+      "epoch": 0.6001419849043435,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 11769
+    },
+    {
+      "epoch": 0.6001929783604488,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6002439718165542,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 11771
+    },
+    {
+      "epoch": 0.6002949652726596,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 11772
+    },
+    {
+      "epoch": 0.6003459587287651,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 11773
+    },
+    {
+      "epoch": 0.6003969521848704,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7742,
+      "step": 11774
+    },
+    {
+      "epoch": 0.6004479456409758,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 11775
+    },
+    {
+      "epoch": 0.6004989390970812,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 11776
+    },
+    {
+      "epoch": 0.6005499325531866,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 11777
+    },
+    {
+      "epoch": 0.6006009260092919,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 11778
+    },
+    {
+      "epoch": 0.6006519194653973,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7734,
+      "step": 11779
+    },
+    {
+      "epoch": 0.6007029129215028,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6007539063776082,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 11781
+    },
+    {
+      "epoch": 0.6008048998337135,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 11782
+    },
+    {
+      "epoch": 0.6008558932898189,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 11783
+    },
+    {
+      "epoch": 0.6009068867459243,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 11784
+    },
+    {
+      "epoch": 0.6009578802020297,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 11785
+    },
+    {
+      "epoch": 0.601008873658135,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 11786
+    },
+    {
+      "epoch": 0.6010598671142405,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 11787
+    },
+    {
+      "epoch": 0.6011108605703459,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7712,
+      "step": 11788
+    },
+    {
+      "epoch": 0.6011618540264513,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 11789
+    },
+    {
+      "epoch": 0.6012128474825567,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 11790
+    },
+    {
+      "epoch": 0.601263840938662,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 11791
+    },
+    {
+      "epoch": 0.6013148343947674,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 11792
+    },
+    {
+      "epoch": 0.6013658278508728,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 11793
+    },
+    {
+      "epoch": 0.6014168213069783,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 11794
+    },
+    {
+      "epoch": 0.6014678147630836,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 11795
+    },
+    {
+      "epoch": 0.601518808219189,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 11796
+    },
+    {
+      "epoch": 0.6015698016752944,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7702,
+      "step": 11797
+    },
+    {
+      "epoch": 0.6016207951313998,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 11798
+    },
+    {
+      "epoch": 0.6016717885875051,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7872,
+      "step": 11799
+    },
+    {
+      "epoch": 0.6017227820436105,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 11800
+    },
+    {
+      "epoch": 0.601773775499716,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 11801
+    },
+    {
+      "epoch": 0.6018247689558214,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 11802
+    },
+    {
+      "epoch": 0.6018757624119268,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 11803
+    },
+    {
+      "epoch": 0.6019267558680321,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 11804
+    },
+    {
+      "epoch": 0.6019777493241375,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 11805
+    },
+    {
+      "epoch": 0.6020287427802429,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 11806
+    },
+    {
+      "epoch": 0.6020797362363483,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 11807
+    },
+    {
+      "epoch": 0.6021307296924537,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 11808
+    },
+    {
+      "epoch": 0.6021817231485591,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 11809
+    },
+    {
+      "epoch": 0.6022327166046645,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6022837100607699,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 11811
+    },
+    {
+      "epoch": 0.6023347035168752,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 11812
+    },
+    {
+      "epoch": 0.6023856969729806,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8311,
+      "step": 11813
+    },
+    {
+      "epoch": 0.602436690429086,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 11814
+    },
+    {
+      "epoch": 0.6024876838851915,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 11815
+    },
+    {
+      "epoch": 0.6025386773412968,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 11816
+    },
+    {
+      "epoch": 0.6025896707974022,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11817
+    },
+    {
+      "epoch": 0.6026406642535076,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7753,
+      "step": 11818
+    },
+    {
+      "epoch": 0.602691657709613,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 11819
+    },
+    {
+      "epoch": 0.6027426511657183,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6027936446218237,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 11821
+    },
+    {
+      "epoch": 0.6028446380779292,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 11822
+    },
+    {
+      "epoch": 0.6028956315340346,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 11823
+    },
+    {
+      "epoch": 0.60294662499014,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 11824
+    },
+    {
+      "epoch": 0.6029976184462453,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 11825
+    },
+    {
+      "epoch": 0.6030486119023507,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 11826
+    },
+    {
+      "epoch": 0.6030996053584561,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8404,
+      "step": 11827
+    },
+    {
+      "epoch": 0.6031505988145615,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 11828
+    },
+    {
+      "epoch": 0.6032015922706669,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 11829
+    },
+    {
+      "epoch": 0.6032525857267723,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6033035791828777,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 11831
+    },
+    {
+      "epoch": 0.6033545726389831,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 11832
+    },
+    {
+      "epoch": 0.6034055660950884,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 11833
+    },
+    {
+      "epoch": 0.6034565595511938,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 11834
+    },
+    {
+      "epoch": 0.6035075530072992,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 11835
+    },
+    {
+      "epoch": 0.6035585464634047,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7719,
+      "step": 11836
+    },
+    {
+      "epoch": 0.60360953991951,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 11837
+    },
+    {
+      "epoch": 0.6036605333756154,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 11838
+    },
+    {
+      "epoch": 0.6037115268317208,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 11839
+    },
+    {
+      "epoch": 0.6037625202878262,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6038135137439316,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 11841
+    },
+    {
+      "epoch": 0.6038645072000369,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7784,
+      "step": 11842
+    },
+    {
+      "epoch": 0.6039155006561424,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 11843
+    },
+    {
+      "epoch": 0.6039664941122478,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 11844
+    },
+    {
+      "epoch": 0.6040174875683532,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 11845
+    },
+    {
+      "epoch": 0.6040684810244585,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 11846
+    },
+    {
+      "epoch": 0.6041194744805639,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 11847
+    },
+    {
+      "epoch": 0.6041704679366693,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 11848
+    },
+    {
+      "epoch": 0.6042214613927747,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 11849
+    },
+    {
+      "epoch": 0.6042724548488801,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6043234483049855,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 11851
+    },
+    {
+      "epoch": 0.6043744417610909,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 11852
+    },
+    {
+      "epoch": 0.6044254352171963,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 11853
+    },
+    {
+      "epoch": 0.6044764286733016,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7788,
+      "step": 11854
+    },
+    {
+      "epoch": 0.604527422129407,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 11855
+    },
+    {
+      "epoch": 0.6045784155855124,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 11856
+    },
+    {
+      "epoch": 0.6046294090416179,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7633,
+      "step": 11857
+    },
+    {
+      "epoch": 0.6046804024977233,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 11858
+    },
+    {
+      "epoch": 0.6047313959538286,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 11859
+    },
+    {
+      "epoch": 0.604782389409934,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6048333828660394,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 11861
+    },
+    {
+      "epoch": 0.6048843763221448,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7798,
+      "step": 11862
+    },
+    {
+      "epoch": 0.6049353697782501,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 11863
+    },
+    {
+      "epoch": 0.6049863632343556,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 11864
+    },
+    {
+      "epoch": 0.605037356690461,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8358,
+      "step": 11865
+    },
+    {
+      "epoch": 0.6050883501465664,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 11866
+    },
+    {
+      "epoch": 0.6051393436026717,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8267,
+      "step": 11867
+    },
+    {
+      "epoch": 0.6051903370587771,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 11868
+    },
+    {
+      "epoch": 0.6052413305148825,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 11869
+    },
+    {
+      "epoch": 0.6052923239709879,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6053433174270934,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 11871
+    },
+    {
+      "epoch": 0.6053943108831987,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 11872
+    },
+    {
+      "epoch": 0.6054453043393041,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 11873
+    },
+    {
+      "epoch": 0.6054962977954095,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7644,
+      "step": 11874
+    },
+    {
+      "epoch": 0.6055472912515149,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 11875
+    },
+    {
+      "epoch": 0.6055982847076202,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 11876
+    },
+    {
+      "epoch": 0.6056492781637256,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 11877
+    },
+    {
+      "epoch": 0.6057002716198311,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 11878
+    },
+    {
+      "epoch": 0.6057512650759365,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 11879
+    },
+    {
+      "epoch": 0.6058022585320418,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6058532519881472,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 11881
+    },
+    {
+      "epoch": 0.6059042454442526,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 11882
+    },
+    {
+      "epoch": 0.605955238900358,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 11883
+    },
+    {
+      "epoch": 0.6060062323564633,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 11884
+    },
+    {
+      "epoch": 0.6060572258125688,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 11885
+    },
+    {
+      "epoch": 0.6061082192686742,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7743,
+      "step": 11886
+    },
+    {
+      "epoch": 0.6061592127247796,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 11887
+    },
+    {
+      "epoch": 0.606210206180885,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 11888
+    },
+    {
+      "epoch": 0.6062611996369903,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7721,
+      "step": 11889
+    },
+    {
+      "epoch": 0.6063121930930957,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7746,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6063631865492011,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7629,
+      "step": 11891
+    },
+    {
+      "epoch": 0.6064141800053066,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 11892
+    },
+    {
+      "epoch": 0.6064651734614119,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 11893
+    },
+    {
+      "epoch": 0.6065161669175173,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 11894
+    },
+    {
+      "epoch": 0.6065671603736227,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.778,
+      "step": 11895
+    },
+    {
+      "epoch": 0.6066181538297281,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 11896
+    },
+    {
+      "epoch": 0.6066691472858334,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 11897
+    },
+    {
+      "epoch": 0.6067201407419388,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 11898
+    },
+    {
+      "epoch": 0.6067711341980443,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 11899
+    },
+    {
+      "epoch": 0.6068221276541497,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7829,
+      "step": 11900
+    },
+    {
+      "epoch": 0.606873121110255,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 11901
+    },
+    {
+      "epoch": 0.6069241145663604,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 11902
+    },
+    {
+      "epoch": 0.6069751080224658,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 11903
+    },
+    {
+      "epoch": 0.6070261014785712,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 11904
+    },
+    {
+      "epoch": 0.6070770949346765,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 11905
+    },
+    {
+      "epoch": 0.607128088390782,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 11906
+    },
+    {
+      "epoch": 0.6071790818468874,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 11907
+    },
+    {
+      "epoch": 0.6072300753029928,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 11908
+    },
+    {
+      "epoch": 0.6072810687590982,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7675,
+      "step": 11909
+    },
+    {
+      "epoch": 0.6073320622152035,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 11910
+    },
+    {
+      "epoch": 0.6073830556713089,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 11911
+    },
+    {
+      "epoch": 0.6074340491274143,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 11912
+    },
+    {
+      "epoch": 0.6074850425835198,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7734,
+      "step": 11913
+    },
+    {
+      "epoch": 0.6075360360396251,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 11914
+    },
+    {
+      "epoch": 0.6075870294957305,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 11915
+    },
+    {
+      "epoch": 0.6076380229518359,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 11916
+    },
+    {
+      "epoch": 0.6076890164079413,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 11917
+    },
+    {
+      "epoch": 0.6077400098640466,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 11918
+    },
+    {
+      "epoch": 0.607791003320152,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 11919
+    },
+    {
+      "epoch": 0.6078419967762575,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 11920
+    },
+    {
+      "epoch": 0.6078929902323629,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 11921
+    },
+    {
+      "epoch": 0.6079439836884682,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7712,
+      "step": 11922
+    },
+    {
+      "epoch": 0.6079949771445736,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 11923
+    },
+    {
+      "epoch": 0.608045970600679,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 11924
+    },
+    {
+      "epoch": 0.6080969640567844,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7604,
+      "step": 11925
+    },
+    {
+      "epoch": 0.6081479575128897,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 11926
+    },
+    {
+      "epoch": 0.6081989509689952,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 11927
+    },
+    {
+      "epoch": 0.6082499444251006,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 11928
+    },
+    {
+      "epoch": 0.608300937881206,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 11929
+    },
+    {
+      "epoch": 0.6083519313373114,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 11930
+    },
+    {
+      "epoch": 0.6084029247934167,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 11931
+    },
+    {
+      "epoch": 0.6084539182495221,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 11932
+    },
+    {
+      "epoch": 0.6085049117056275,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 11933
+    },
+    {
+      "epoch": 0.608555905161733,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 11934
+    },
+    {
+      "epoch": 0.6086068986178383,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 11935
+    },
+    {
+      "epoch": 0.6086578920739437,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 11936
+    },
+    {
+      "epoch": 0.6087088855300491,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 11937
+    },
+    {
+      "epoch": 0.6087598789861545,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7644,
+      "step": 11938
+    },
+    {
+      "epoch": 0.6088108724422598,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 11939
+    },
+    {
+      "epoch": 0.6088618658983652,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7677,
+      "step": 11940
+    },
+    {
+      "epoch": 0.6089128593544707,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 11941
+    },
+    {
+      "epoch": 0.6089638528105761,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 11942
+    },
+    {
+      "epoch": 0.6090148462666815,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 11943
+    },
+    {
+      "epoch": 0.6090658397227868,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 11944
+    },
+    {
+      "epoch": 0.6091168331788922,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 11945
+    },
+    {
+      "epoch": 0.6091678266349976,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 11946
+    },
+    {
+      "epoch": 0.609218820091103,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 11947
+    },
+    {
+      "epoch": 0.6092698135472084,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7717,
+      "step": 11948
+    },
+    {
+      "epoch": 0.6093208070033138,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8424,
+      "step": 11949
+    },
+    {
+      "epoch": 0.6093718004594192,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 11950
+    },
+    {
+      "epoch": 0.6094227939155246,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7706,
+      "step": 11951
+    },
+    {
+      "epoch": 0.6094737873716299,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 11952
+    },
+    {
+      "epoch": 0.6095247808277353,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 11953
+    },
+    {
+      "epoch": 0.6095757742838407,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 11954
+    },
+    {
+      "epoch": 0.6096267677399462,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 11955
+    },
+    {
+      "epoch": 0.6096777611960515,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7772,
+      "step": 11956
+    },
+    {
+      "epoch": 0.6097287546521569,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 11957
+    },
+    {
+      "epoch": 0.6097797481082623,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 11958
+    },
+    {
+      "epoch": 0.6098307415643677,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 11959
+    },
+    {
+      "epoch": 0.609881735020473,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 11960
+    },
+    {
+      "epoch": 0.6099327284765784,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 11961
+    },
+    {
+      "epoch": 0.6099837219326839,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 11962
+    },
+    {
+      "epoch": 0.6100347153887893,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 11963
+    },
+    {
+      "epoch": 0.6100857088448947,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 11964
+    },
+    {
+      "epoch": 0.610136702301,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 11965
+    },
+    {
+      "epoch": 0.6101876957571054,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 11966
+    },
+    {
+      "epoch": 0.6102386892132108,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 11967
+    },
+    {
+      "epoch": 0.6102896826693162,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 11968
+    },
+    {
+      "epoch": 0.6103406761254216,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 11969
+    },
+    {
+      "epoch": 0.610391669581527,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 11970
+    },
+    {
+      "epoch": 0.6104426630376324,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 11971
+    },
+    {
+      "epoch": 0.6104936564937378,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 11972
+    },
+    {
+      "epoch": 0.6105446499498431,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 11973
+    },
+    {
+      "epoch": 0.6105956434059485,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 11974
+    },
+    {
+      "epoch": 0.6106466368620539,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 11975
+    },
+    {
+      "epoch": 0.6106976303181594,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 11976
+    },
+    {
+      "epoch": 0.6107486237742648,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 11977
+    },
+    {
+      "epoch": 0.6107996172303701,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 11978
+    },
+    {
+      "epoch": 0.6108506106864755,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7813,
+      "step": 11979
+    },
+    {
+      "epoch": 0.6109016041425809,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 11980
+    },
+    {
+      "epoch": 0.6109525975986863,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 11981
+    },
+    {
+      "epoch": 0.6110035910547916,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 11982
+    },
+    {
+      "epoch": 0.6110545845108971,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 11983
+    },
+    {
+      "epoch": 0.6111055779670025,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 11984
+    },
+    {
+      "epoch": 0.6111565714231079,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 11985
+    },
+    {
+      "epoch": 0.6112075648792132,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 11986
+    },
+    {
+      "epoch": 0.6112585583353186,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 11987
+    },
+    {
+      "epoch": 0.611309551791424,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 11988
+    },
+    {
+      "epoch": 0.6113605452475294,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 11989
+    },
+    {
+      "epoch": 0.6114115387036348,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 11990
+    },
+    {
+      "epoch": 0.6114625321597402,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 11991
+    },
+    {
+      "epoch": 0.6115135256158456,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.766,
+      "step": 11992
+    },
+    {
+      "epoch": 0.611564519071951,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 11993
+    },
+    {
+      "epoch": 0.6116155125280563,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 11994
+    },
+    {
+      "epoch": 0.6116665059841617,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8344,
+      "step": 11995
+    },
+    {
+      "epoch": 0.6117174994402671,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 11996
+    },
+    {
+      "epoch": 0.6117684928963726,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 11997
+    },
+    {
+      "epoch": 0.611819486352478,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 11998
+    },
+    {
+      "epoch": 0.6118704798085833,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 11999
+    },
+    {
+      "epoch": 0.6119214732646887,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6119724667207941,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 12001
+    },
+    {
+      "epoch": 0.6120234601768995,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 12002
+    },
+    {
+      "epoch": 0.6120744536330048,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 12003
+    },
+    {
+      "epoch": 0.6121254470891103,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 12004
+    },
+    {
+      "epoch": 0.6121764405452157,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7719,
+      "step": 12005
+    },
+    {
+      "epoch": 0.6122274340013211,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 12006
+    },
+    {
+      "epoch": 0.6122784274574264,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12007
+    },
+    {
+      "epoch": 0.6123294209135318,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 12008
+    },
+    {
+      "epoch": 0.6123804143696372,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 12009
+    },
+    {
+      "epoch": 0.6124314078257426,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 12010
+    },
+    {
+      "epoch": 0.612482401281848,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 12011
+    },
+    {
+      "epoch": 0.6125333947379534,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 12012
+    },
+    {
+      "epoch": 0.6125843881940588,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 12013
+    },
+    {
+      "epoch": 0.6126353816501642,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 12014
+    },
+    {
+      "epoch": 0.6126863751062696,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 12015
+    },
+    {
+      "epoch": 0.6127373685623749,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 12016
+    },
+    {
+      "epoch": 0.6127883620184803,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 12017
+    },
+    {
+      "epoch": 0.6128393554745858,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 12018
+    },
+    {
+      "epoch": 0.6128903489306912,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 12019
+    },
+    {
+      "epoch": 0.6129413423867965,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 12020
+    },
+    {
+      "epoch": 0.6129923358429019,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12021
+    },
+    {
+      "epoch": 0.6130433292990073,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 12022
+    },
+    {
+      "epoch": 0.6130943227551127,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12023
+    },
+    {
+      "epoch": 0.613145316211218,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 12024
+    },
+    {
+      "epoch": 0.6131963096673235,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 12025
+    },
+    {
+      "epoch": 0.6132473031234289,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 12026
+    },
+    {
+      "epoch": 0.6132982965795343,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 12027
+    },
+    {
+      "epoch": 0.6133492900356396,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 12028
+    },
+    {
+      "epoch": 0.613400283491745,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8383,
+      "step": 12029
+    },
+    {
+      "epoch": 0.6134512769478504,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 12030
+    },
+    {
+      "epoch": 0.6135022704039558,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 12031
+    },
+    {
+      "epoch": 0.6135532638600613,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 12032
+    },
+    {
+      "epoch": 0.6136042573161666,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 12033
+    },
+    {
+      "epoch": 0.613655250772272,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 12034
+    },
+    {
+      "epoch": 0.6137062442283774,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 12035
+    },
+    {
+      "epoch": 0.6137572376844828,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 12036
+    },
+    {
+      "epoch": 0.6138082311405881,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 12037
+    },
+    {
+      "epoch": 0.6138592245966935,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 12038
+    },
+    {
+      "epoch": 0.613910218052799,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 12039
+    },
+    {
+      "epoch": 0.6139612115089044,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 12040
+    },
+    {
+      "epoch": 0.6140122049650097,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 12041
+    },
+    {
+      "epoch": 0.6140631984211151,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 12042
+    },
+    {
+      "epoch": 0.6141141918772205,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7627,
+      "step": 12043
+    },
+    {
+      "epoch": 0.6141651853333259,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 12044
+    },
+    {
+      "epoch": 0.6142161787894312,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 12045
+    },
+    {
+      "epoch": 0.6142671722455367,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 12046
+    },
+    {
+      "epoch": 0.6143181657016421,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 12047
+    },
+    {
+      "epoch": 0.6143691591577475,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 12048
+    },
+    {
+      "epoch": 0.6144201526138529,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.788,
+      "step": 12049
+    },
+    {
+      "epoch": 0.6144711460699582,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 12050
+    },
+    {
+      "epoch": 0.6145221395260636,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 12051
+    },
+    {
+      "epoch": 0.614573132982169,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 12052
+    },
+    {
+      "epoch": 0.6146241264382745,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 12053
+    },
+    {
+      "epoch": 0.6146751198943798,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 12054
+    },
+    {
+      "epoch": 0.6147261133504852,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 12055
+    },
+    {
+      "epoch": 0.6147771068065906,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 12056
+    },
+    {
+      "epoch": 0.614828100262696,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 12057
+    },
+    {
+      "epoch": 0.6148790937188013,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 12058
+    },
+    {
+      "epoch": 0.6149300871749067,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 12059
+    },
+    {
+      "epoch": 0.6149810806310122,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 12060
+    },
+    {
+      "epoch": 0.6150320740871176,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 12061
+    },
+    {
+      "epoch": 0.615083067543223,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7669,
+      "step": 12062
+    },
+    {
+      "epoch": 0.6151340609993283,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 12063
+    },
+    {
+      "epoch": 0.6151850544554337,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 12064
+    },
+    {
+      "epoch": 0.6152360479115391,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 12065
+    },
+    {
+      "epoch": 0.6152870413676444,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 12066
+    },
+    {
+      "epoch": 0.6153380348237499,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 12067
+    },
+    {
+      "epoch": 0.6153890282798553,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 12068
+    },
+    {
+      "epoch": 0.6154400217359607,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12069
+    },
+    {
+      "epoch": 0.6154910151920661,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 12070
+    },
+    {
+      "epoch": 0.6155420086481714,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 12071
+    },
+    {
+      "epoch": 0.6155930021042768,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 12072
+    },
+    {
+      "epoch": 0.6156439955603822,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 12073
+    },
+    {
+      "epoch": 0.6156949890164877,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12074
+    },
+    {
+      "epoch": 0.615745982472593,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 12075
+    },
+    {
+      "epoch": 0.6157969759286984,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 12076
+    },
+    {
+      "epoch": 0.6158479693848038,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 12077
+    },
+    {
+      "epoch": 0.6158989628409092,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 12078
+    },
+    {
+      "epoch": 0.6159499562970145,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 12079
+    },
+    {
+      "epoch": 0.6160009497531199,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7815,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6160519432092254,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 12081
+    },
+    {
+      "epoch": 0.6161029366653308,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 12082
+    },
+    {
+      "epoch": 0.6161539301214362,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 12083
+    },
+    {
+      "epoch": 0.6162049235775415,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12084
+    },
+    {
+      "epoch": 0.6162559170336469,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 12085
+    },
+    {
+      "epoch": 0.6163069104897523,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7753,
+      "step": 12086
+    },
+    {
+      "epoch": 0.6163579039458577,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 12087
+    },
+    {
+      "epoch": 0.6164088974019631,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.788,
+      "step": 12088
+    },
+    {
+      "epoch": 0.6164598908580685,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 12089
+    },
+    {
+      "epoch": 0.6165108843141739,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6165618777702793,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 12091
+    },
+    {
+      "epoch": 0.6166128712263846,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 12092
+    },
+    {
+      "epoch": 0.61666386468249,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 12093
+    },
+    {
+      "epoch": 0.6167148581385954,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 12094
+    },
+    {
+      "epoch": 0.6167658515947009,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12095
+    },
+    {
+      "epoch": 0.6168168450508063,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 12096
+    },
+    {
+      "epoch": 0.6168678385069116,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 12097
+    },
+    {
+      "epoch": 0.616918831963017,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7556,
+      "step": 12098
+    },
+    {
+      "epoch": 0.6169698254191224,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 12099
+    },
+    {
+      "epoch": 0.6170208188752278,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6170718123313331,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 12101
+    },
+    {
+      "epoch": 0.6171228057874386,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 12102
+    },
+    {
+      "epoch": 0.617173799243544,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 12103
+    },
+    {
+      "epoch": 0.6172247926996494,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 12104
+    },
+    {
+      "epoch": 0.6172757861557547,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 12105
+    },
+    {
+      "epoch": 0.6173267796118601,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 12106
+    },
+    {
+      "epoch": 0.6173777730679655,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 12107
+    },
+    {
+      "epoch": 0.6174287665240709,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 12108
+    },
+    {
+      "epoch": 0.6174797599801763,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 12109
+    },
+    {
+      "epoch": 0.6175307534362817,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6175817468923871,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12111
+    },
+    {
+      "epoch": 0.6176327403484925,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7589,
+      "step": 12112
+    },
+    {
+      "epoch": 0.6176837338045978,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 12113
+    },
+    {
+      "epoch": 0.6177347272607032,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 12114
+    },
+    {
+      "epoch": 0.6177857207168086,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 12115
+    },
+    {
+      "epoch": 0.6178367141729141,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 12116
+    },
+    {
+      "epoch": 0.6178877076290195,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 12117
+    },
+    {
+      "epoch": 0.6179387010851248,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 12118
+    },
+    {
+      "epoch": 0.6179896945412302,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 12119
+    },
+    {
+      "epoch": 0.6180406879973356,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 12120
+    },
+    {
+      "epoch": 0.618091681453441,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 12121
+    },
+    {
+      "epoch": 0.6181426749095463,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 12122
+    },
+    {
+      "epoch": 0.6181936683656518,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 12123
+    },
+    {
+      "epoch": 0.6182446618217572,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 12124
+    },
+    {
+      "epoch": 0.6182956552778626,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 12125
+    },
+    {
+      "epoch": 0.6183466487339679,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7766,
+      "step": 12126
+    },
+    {
+      "epoch": 0.6183976421900733,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12127
+    },
+    {
+      "epoch": 0.6184486356461787,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7754,
+      "step": 12128
+    },
+    {
+      "epoch": 0.6184996291022841,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12129
+    },
+    {
+      "epoch": 0.6185506225583896,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6186016160144949,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 12131
+    },
+    {
+      "epoch": 0.6186526094706003,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 12132
+    },
+    {
+      "epoch": 0.6187036029267057,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 12133
+    },
+    {
+      "epoch": 0.618754596382811,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7686,
+      "step": 12134
+    },
+    {
+      "epoch": 0.6188055898389164,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7758,
+      "step": 12135
+    },
+    {
+      "epoch": 0.6188565832950218,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 12136
+    },
+    {
+      "epoch": 0.6189075767511273,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 12137
+    },
+    {
+      "epoch": 0.6189585702072327,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12138
+    },
+    {
+      "epoch": 0.619009563663338,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 12139
+    },
+    {
+      "epoch": 0.6190605571194434,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7563,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6191115505755488,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7736,
+      "step": 12141
+    },
+    {
+      "epoch": 0.6191625440316542,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 12142
+    },
+    {
+      "epoch": 0.6192135374877595,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 12143
+    },
+    {
+      "epoch": 0.619264530943865,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 12144
+    },
+    {
+      "epoch": 0.6193155243999704,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 12145
+    },
+    {
+      "epoch": 0.6193665178560758,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 12146
+    },
+    {
+      "epoch": 0.6194175113121811,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7626,
+      "step": 12147
+    },
+    {
+      "epoch": 0.6194685047682865,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 12148
+    },
+    {
+      "epoch": 0.6195194982243919,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 12149
+    },
+    {
+      "epoch": 0.6195704916804973,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 12150
+    },
+    {
+      "epoch": 0.6196214851366028,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7789,
+      "step": 12151
+    },
+    {
+      "epoch": 0.6196724785927081,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 12152
+    },
+    {
+      "epoch": 0.6197234720488135,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7713,
+      "step": 12153
+    },
+    {
+      "epoch": 0.6197744655049189,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12154
+    },
+    {
+      "epoch": 0.6198254589610243,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 12155
+    },
+    {
+      "epoch": 0.6198764524171296,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 12156
+    },
+    {
+      "epoch": 0.619927445873235,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7801,
+      "step": 12157
+    },
+    {
+      "epoch": 0.6199784393293405,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7666,
+      "step": 12158
+    },
+    {
+      "epoch": 0.6200294327854459,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 12159
+    },
+    {
+      "epoch": 0.6200804262415512,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6201314196976566,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7772,
+      "step": 12161
+    },
+    {
+      "epoch": 0.620182413153762,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 12162
+    },
+    {
+      "epoch": 0.6202334066098674,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 12163
+    },
+    {
+      "epoch": 0.6202844000659727,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 12164
+    },
+    {
+      "epoch": 0.6203353935220782,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 12165
+    },
+    {
+      "epoch": 0.6203863869781836,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 12166
+    },
+    {
+      "epoch": 0.620437380434289,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 12167
+    },
+    {
+      "epoch": 0.6204883738903944,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 12168
+    },
+    {
+      "epoch": 0.6205393673464997,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 12169
+    },
+    {
+      "epoch": 0.6205903608026051,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6206413542587105,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 12171
+    },
+    {
+      "epoch": 0.620692347714816,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7664,
+      "step": 12172
+    },
+    {
+      "epoch": 0.6207433411709213,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 12173
+    },
+    {
+      "epoch": 0.6207943346270267,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 12174
+    },
+    {
+      "epoch": 0.6208453280831321,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 12175
+    },
+    {
+      "epoch": 0.6208963215392375,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7739,
+      "step": 12176
+    },
+    {
+      "epoch": 0.6209473149953428,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 12177
+    },
+    {
+      "epoch": 0.6209983084514482,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 12178
+    },
+    {
+      "epoch": 0.6210493019075537,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 12179
+    },
+    {
+      "epoch": 0.6211002953636591,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 12180
+    },
+    {
+      "epoch": 0.6211512888197644,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 12181
+    },
+    {
+      "epoch": 0.6212022822758698,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 12182
+    },
+    {
+      "epoch": 0.6212532757319752,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 12183
+    },
+    {
+      "epoch": 0.6213042691880806,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 12184
+    },
+    {
+      "epoch": 0.621355262644186,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 12185
+    },
+    {
+      "epoch": 0.6214062561002914,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 12186
+    },
+    {
+      "epoch": 0.6214572495563968,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7724,
+      "step": 12187
+    },
+    {
+      "epoch": 0.6215082430125022,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 12188
+    },
+    {
+      "epoch": 0.6215592364686076,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7772,
+      "step": 12189
+    },
+    {
+      "epoch": 0.6216102299247129,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 12190
+    },
+    {
+      "epoch": 0.6216612233808183,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 12191
+    },
+    {
+      "epoch": 0.6217122168369237,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 12192
+    },
+    {
+      "epoch": 0.6217632102930292,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7828,
+      "step": 12193
+    },
+    {
+      "epoch": 0.6218142037491345,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 12194
+    },
+    {
+      "epoch": 0.6218651972052399,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 12195
+    },
+    {
+      "epoch": 0.6219161906613453,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12196
+    },
+    {
+      "epoch": 0.6219671841174507,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 12197
+    },
+    {
+      "epoch": 0.622018177573556,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7701,
+      "step": 12198
+    },
+    {
+      "epoch": 0.6220691710296614,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 12199
+    },
+    {
+      "epoch": 0.6221201644857669,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6221711579418723,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 12201
+    },
+    {
+      "epoch": 0.6222221513979777,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 12202
+    },
+    {
+      "epoch": 0.622273144854083,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 12203
+    },
+    {
+      "epoch": 0.6223241383101884,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 12204
+    },
+    {
+      "epoch": 0.6223751317662938,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 12205
+    },
+    {
+      "epoch": 0.6224261252223992,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 12206
+    },
+    {
+      "epoch": 0.6224771186785046,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 12207
+    },
+    {
+      "epoch": 0.62252811213461,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 12208
+    },
+    {
+      "epoch": 0.6225791055907154,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 12209
+    },
+    {
+      "epoch": 0.6226300990468208,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7787,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6226810925029261,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 12211
+    },
+    {
+      "epoch": 0.6227320859590315,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 12212
+    },
+    {
+      "epoch": 0.6227830794151369,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 12213
+    },
+    {
+      "epoch": 0.6228340728712424,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 12214
+    },
+    {
+      "epoch": 0.6228850663273477,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 12215
+    },
+    {
+      "epoch": 0.6229360597834531,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7726,
+      "step": 12216
+    },
+    {
+      "epoch": 0.6229870532395585,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 12217
+    },
+    {
+      "epoch": 0.6230380466956639,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 12218
+    },
+    {
+      "epoch": 0.6230890401517692,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 12219
+    },
+    {
+      "epoch": 0.6231400336078746,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6231910270639801,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 12221
+    },
+    {
+      "epoch": 0.6232420205200855,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 12222
+    },
+    {
+      "epoch": 0.6232930139761909,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 12223
+    },
+    {
+      "epoch": 0.6233440074322962,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 12224
+    },
+    {
+      "epoch": 0.6233950008884016,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 12225
+    },
+    {
+      "epoch": 0.623445994344507,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 12226
+    },
+    {
+      "epoch": 0.6234969878006124,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 12227
+    },
+    {
+      "epoch": 0.6235479812567178,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 12228
+    },
+    {
+      "epoch": 0.6235989747128232,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 12229
+    },
+    {
+      "epoch": 0.6236499681689286,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 12230
+    },
+    {
+      "epoch": 0.623700961625034,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 12231
+    },
+    {
+      "epoch": 0.6237519550811393,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 12232
+    },
+    {
+      "epoch": 0.6238029485372447,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7788,
+      "step": 12233
+    },
+    {
+      "epoch": 0.6238539419933501,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 12234
+    },
+    {
+      "epoch": 0.6239049354494556,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 12235
+    },
+    {
+      "epoch": 0.623955928905561,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 12236
+    },
+    {
+      "epoch": 0.6240069223616663,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7548,
+      "step": 12237
+    },
+    {
+      "epoch": 0.6240579158177717,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 12238
+    },
+    {
+      "epoch": 0.6241089092738771,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 12239
+    },
+    {
+      "epoch": 0.6241599027299825,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6242108961860878,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7814,
+      "step": 12241
+    },
+    {
+      "epoch": 0.6242618896421933,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 12242
+    },
+    {
+      "epoch": 0.6243128830982987,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 12243
+    },
+    {
+      "epoch": 0.6243638765544041,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 12244
+    },
+    {
+      "epoch": 0.6244148700105094,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7708,
+      "step": 12245
+    },
+    {
+      "epoch": 0.6244658634666148,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 12246
+    },
+    {
+      "epoch": 0.6245168569227202,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7733,
+      "step": 12247
+    },
+    {
+      "epoch": 0.6245678503788256,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 12248
+    },
+    {
+      "epoch": 0.624618843834931,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 12249
+    },
+    {
+      "epoch": 0.6246698372910364,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6247208307471418,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 12251
+    },
+    {
+      "epoch": 0.6247718242032472,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12252
+    },
+    {
+      "epoch": 0.6248228176593525,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 12253
+    },
+    {
+      "epoch": 0.6248738111154579,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 12254
+    },
+    {
+      "epoch": 0.6249248045715633,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7776,
+      "step": 12255
+    },
+    {
+      "epoch": 0.6249757980276688,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 12256
+    },
+    {
+      "epoch": 0.6250267914837742,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 12257
+    },
+    {
+      "epoch": 0.6250777849398795,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 12258
+    },
+    {
+      "epoch": 0.6251287783959849,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 12259
+    },
+    {
+      "epoch": 0.6251797718520903,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6252307653081957,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 12261
+    },
+    {
+      "epoch": 0.625281758764301,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12262
+    },
+    {
+      "epoch": 0.6253327522204065,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 12263
+    },
+    {
+      "epoch": 0.6253837456765119,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 12264
+    },
+    {
+      "epoch": 0.6254347391326173,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12265
+    },
+    {
+      "epoch": 0.6254857325887226,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 12266
+    },
+    {
+      "epoch": 0.625536726044828,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 12267
+    },
+    {
+      "epoch": 0.6255877195009334,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 12268
+    },
+    {
+      "epoch": 0.6256387129570388,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 12269
+    },
+    {
+      "epoch": 0.6256897064131443,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6257406998692496,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 12271
+    },
+    {
+      "epoch": 0.625791693325355,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 12272
+    },
+    {
+      "epoch": 0.6258426867814604,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 12273
+    },
+    {
+      "epoch": 0.6258936802375658,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 12274
+    },
+    {
+      "epoch": 0.6259446736936711,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 12275
+    },
+    {
+      "epoch": 0.6259956671497765,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 12276
+    },
+    {
+      "epoch": 0.626046660605882,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 12277
+    },
+    {
+      "epoch": 0.6260976540619874,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12278
+    },
+    {
+      "epoch": 0.6261486475180927,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 12279
+    },
+    {
+      "epoch": 0.6261996409741981,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6262506344303035,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7669,
+      "step": 12281
+    },
+    {
+      "epoch": 0.6263016278864089,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7665,
+      "step": 12282
+    },
+    {
+      "epoch": 0.6263526213425142,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 12283
+    },
+    {
+      "epoch": 0.6264036147986197,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 12284
+    },
+    {
+      "epoch": 0.6264546082547251,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 12285
+    },
+    {
+      "epoch": 0.6265056017108305,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12286
+    },
+    {
+      "epoch": 0.6265565951669358,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 12287
+    },
+    {
+      "epoch": 0.6266075886230412,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 12288
+    },
+    {
+      "epoch": 0.6266585820791466,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 12289
+    },
+    {
+      "epoch": 0.626709575535252,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6267605689913575,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 12291
+    },
+    {
+      "epoch": 0.6268115624474628,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8385,
+      "step": 12292
+    },
+    {
+      "epoch": 0.6268625559035682,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 12293
+    },
+    {
+      "epoch": 0.6269135493596736,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12294
+    },
+    {
+      "epoch": 0.626964542815779,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12295
+    },
+    {
+      "epoch": 0.6270155362718843,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 12296
+    },
+    {
+      "epoch": 0.6270665297279897,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 12297
+    },
+    {
+      "epoch": 0.6271175231840952,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 12298
+    },
+    {
+      "epoch": 0.6271685166402006,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 12299
+    },
+    {
+      "epoch": 0.6272195100963059,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6272705035524113,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 12301
+    },
+    {
+      "epoch": 0.6273214970085167,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7635,
+      "step": 12302
+    },
+    {
+      "epoch": 0.6273724904646221,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 12303
+    },
+    {
+      "epoch": 0.6274234839207274,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7815,
+      "step": 12304
+    },
+    {
+      "epoch": 0.6274744773768329,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 12305
+    },
+    {
+      "epoch": 0.6275254708329383,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 12306
+    },
+    {
+      "epoch": 0.6275764642890437,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 12307
+    },
+    {
+      "epoch": 0.627627457745149,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 12308
+    },
+    {
+      "epoch": 0.6276784512012544,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 12309
+    },
+    {
+      "epoch": 0.6277294446573598,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 12310
+    },
+    {
+      "epoch": 0.6277804381134652,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 12311
+    },
+    {
+      "epoch": 0.6278314315695707,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 12312
+    },
+    {
+      "epoch": 0.627882425025676,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 12313
+    },
+    {
+      "epoch": 0.6279334184817814,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 12314
+    },
+    {
+      "epoch": 0.6279844119378868,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 12315
+    },
+    {
+      "epoch": 0.6280354053939922,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 12316
+    },
+    {
+      "epoch": 0.6280863988500975,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 12317
+    },
+    {
+      "epoch": 0.6281373923062029,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7759,
+      "step": 12318
+    },
+    {
+      "epoch": 0.6281883857623084,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 12319
+    },
+    {
+      "epoch": 0.6282393792184138,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 12320
+    },
+    {
+      "epoch": 0.6282903726745191,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 12321
+    },
+    {
+      "epoch": 0.6283413661306245,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 12322
+    },
+    {
+      "epoch": 0.6283923595867299,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 12323
+    },
+    {
+      "epoch": 0.6284433530428353,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 12324
+    },
+    {
+      "epoch": 0.6284943464989406,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12325
+    },
+    {
+      "epoch": 0.6285453399550461,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 12326
+    },
+    {
+      "epoch": 0.6285963334111515,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12327
+    },
+    {
+      "epoch": 0.6286473268672569,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 12328
+    },
+    {
+      "epoch": 0.6286983203233623,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7667,
+      "step": 12329
+    },
+    {
+      "epoch": 0.6287493137794676,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 12330
+    },
+    {
+      "epoch": 0.628800307235573,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 12331
+    },
+    {
+      "epoch": 0.6288513006916784,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 12332
+    },
+    {
+      "epoch": 0.6289022941477839,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 12333
+    },
+    {
+      "epoch": 0.6289532876038892,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 12334
+    },
+    {
+      "epoch": 0.6290042810599946,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 12335
+    },
+    {
+      "epoch": 0.6290552745161,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7709,
+      "step": 12336
+    },
+    {
+      "epoch": 0.6291062679722054,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7683,
+      "step": 12337
+    },
+    {
+      "epoch": 0.6291572614283107,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7702,
+      "step": 12338
+    },
+    {
+      "epoch": 0.6292082548844161,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 12339
+    },
+    {
+      "epoch": 0.6292592483405216,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 12340
+    },
+    {
+      "epoch": 0.629310241796627,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 12341
+    },
+    {
+      "epoch": 0.6293612352527324,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 12342
+    },
+    {
+      "epoch": 0.6294122287088377,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 12343
+    },
+    {
+      "epoch": 0.6294632221649431,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 12344
+    },
+    {
+      "epoch": 0.6295142156210485,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12345
+    },
+    {
+      "epoch": 0.6295652090771539,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 12346
+    },
+    {
+      "epoch": 0.6296162025332593,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7828,
+      "step": 12347
+    },
+    {
+      "epoch": 0.6296671959893647,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 12348
+    },
+    {
+      "epoch": 0.6297181894454701,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 12349
+    },
+    {
+      "epoch": 0.6297691829015755,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 12350
+    },
+    {
+      "epoch": 0.6298201763576808,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7523,
+      "step": 12351
+    },
+    {
+      "epoch": 0.6298711698137862,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.777,
+      "step": 12352
+    },
+    {
+      "epoch": 0.6299221632698916,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 12353
+    },
+    {
+      "epoch": 0.6299731567259971,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 12354
+    },
+    {
+      "epoch": 0.6300241501821024,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 12355
+    },
+    {
+      "epoch": 0.6300751436382078,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 12356
+    },
+    {
+      "epoch": 0.6301261370943132,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12357
+    },
+    {
+      "epoch": 0.6301771305504186,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 12358
+    },
+    {
+      "epoch": 0.630228124006524,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7629,
+      "step": 12359
+    },
+    {
+      "epoch": 0.6302791174626293,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6303301109187348,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 12361
+    },
+    {
+      "epoch": 0.6303811043748402,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 12362
+    },
+    {
+      "epoch": 0.6304320978309456,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 12363
+    },
+    {
+      "epoch": 0.6304830912870509,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 12364
+    },
+    {
+      "epoch": 0.6305340847431563,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7704,
+      "step": 12365
+    },
+    {
+      "epoch": 0.6305850781992617,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 12366
+    },
+    {
+      "epoch": 0.6306360716553671,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 12367
+    },
+    {
+      "epoch": 0.6306870651114725,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 12368
+    },
+    {
+      "epoch": 0.6307380585675779,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 12369
+    },
+    {
+      "epoch": 0.6307890520236833,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.766,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6308400454797887,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12371
+    },
+    {
+      "epoch": 0.630891038935894,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 12372
+    },
+    {
+      "epoch": 0.6309420323919994,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7659,
+      "step": 12373
+    },
+    {
+      "epoch": 0.6309930258481048,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 12374
+    },
+    {
+      "epoch": 0.6310440193042103,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 12375
+    },
+    {
+      "epoch": 0.6310950127603157,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 12376
+    },
+    {
+      "epoch": 0.631146006216421,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 12377
+    },
+    {
+      "epoch": 0.6311969996725264,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 12378
+    },
+    {
+      "epoch": 0.6312479931286318,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 12379
+    },
+    {
+      "epoch": 0.6312989865847372,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6313499800408425,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 12381
+    },
+    {
+      "epoch": 0.631400973496948,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 12382
+    },
+    {
+      "epoch": 0.6314519669530534,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 12383
+    },
+    {
+      "epoch": 0.6315029604091588,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7872,
+      "step": 12384
+    },
+    {
+      "epoch": 0.6315539538652641,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 12385
+    },
+    {
+      "epoch": 0.6316049473213695,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 12386
+    },
+    {
+      "epoch": 0.6316559407774749,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 12387
+    },
+    {
+      "epoch": 0.6317069342335803,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 12388
+    },
+    {
+      "epoch": 0.6317579276896857,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 12389
+    },
+    {
+      "epoch": 0.6318089211457911,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6318599146018965,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12391
+    },
+    {
+      "epoch": 0.6319109080580019,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 12392
+    },
+    {
+      "epoch": 0.6319619015141072,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12393
+    },
+    {
+      "epoch": 0.6320128949702126,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7613,
+      "step": 12394
+    },
+    {
+      "epoch": 0.632063888426318,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 12395
+    },
+    {
+      "epoch": 0.6321148818824235,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 12396
+    },
+    {
+      "epoch": 0.6321658753385289,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12397
+    },
+    {
+      "epoch": 0.6322168687946342,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 12398
+    },
+    {
+      "epoch": 0.6322678622507396,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 12399
+    },
+    {
+      "epoch": 0.632318855706845,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6323698491629504,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 12401
+    },
+    {
+      "epoch": 0.6324208426190557,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 12402
+    },
+    {
+      "epoch": 0.6324718360751612,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 12403
+    },
+    {
+      "epoch": 0.6325228295312666,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 12404
+    },
+    {
+      "epoch": 0.632573822987372,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 12405
+    },
+    {
+      "epoch": 0.6326248164434773,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 12406
+    },
+    {
+      "epoch": 0.6326758098995827,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 12407
+    },
+    {
+      "epoch": 0.6327268033556881,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 12408
+    },
+    {
+      "epoch": 0.6327777968117935,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 12409
+    },
+    {
+      "epoch": 0.632828790267899,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6328797837240043,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 12411
+    },
+    {
+      "epoch": 0.6329307771801097,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 12412
+    },
+    {
+      "epoch": 0.6329817706362151,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 12413
+    },
+    {
+      "epoch": 0.6330327640923205,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7733,
+      "step": 12414
+    },
+    {
+      "epoch": 0.6330837575484258,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 12415
+    },
+    {
+      "epoch": 0.6331347510045312,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 12416
+    },
+    {
+      "epoch": 0.6331857444606367,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 12417
+    },
+    {
+      "epoch": 0.6332367379167421,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 12418
+    },
+    {
+      "epoch": 0.6332877313728474,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 12419
+    },
+    {
+      "epoch": 0.6333387248289528,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6333897182850582,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 12421
+    },
+    {
+      "epoch": 0.6334407117411636,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 12422
+    },
+    {
+      "epoch": 0.6334917051972689,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 12423
+    },
+    {
+      "epoch": 0.6335426986533744,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 12424
+    },
+    {
+      "epoch": 0.6335936921094798,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 12425
+    },
+    {
+      "epoch": 0.6336446855655852,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 12426
+    },
+    {
+      "epoch": 0.6336956790216905,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 12427
+    },
+    {
+      "epoch": 0.6337466724777959,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 12428
+    },
+    {
+      "epoch": 0.6337976659339013,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 12429
+    },
+    {
+      "epoch": 0.6338486593900067,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7713,
+      "step": 12430
+    },
+    {
+      "epoch": 0.6338996528461122,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 12431
+    },
+    {
+      "epoch": 0.6339506463022175,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 12432
+    },
+    {
+      "epoch": 0.6340016397583229,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 12433
+    },
+    {
+      "epoch": 0.6340526332144283,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 12434
+    },
+    {
+      "epoch": 0.6341036266705337,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 12435
+    },
+    {
+      "epoch": 0.634154620126639,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 12436
+    },
+    {
+      "epoch": 0.6342056135827444,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 12437
+    },
+    {
+      "epoch": 0.6342566070388499,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 12438
+    },
+    {
+      "epoch": 0.6343076004949553,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 12439
+    },
+    {
+      "epoch": 0.6343585939510606,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 12440
+    },
+    {
+      "epoch": 0.634409587407166,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 12441
+    },
+    {
+      "epoch": 0.6344605808632714,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7777,
+      "step": 12442
+    },
+    {
+      "epoch": 0.6345115743193768,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7775,
+      "step": 12443
+    },
+    {
+      "epoch": 0.6345625677754821,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 12444
+    },
+    {
+      "epoch": 0.6346135612315876,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 12445
+    },
+    {
+      "epoch": 0.634664554687693,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 12446
+    },
+    {
+      "epoch": 0.6347155481437984,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 12447
+    },
+    {
+      "epoch": 0.6347665415999038,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 12448
+    },
+    {
+      "epoch": 0.6348175350560091,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7605,
+      "step": 12449
+    },
+    {
+      "epoch": 0.6348685285121145,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6349195219682199,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7796,
+      "step": 12451
+    },
+    {
+      "epoch": 0.6349705154243254,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 12452
+    },
+    {
+      "epoch": 0.6350215088804307,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7519,
+      "step": 12453
+    },
+    {
+      "epoch": 0.6350725023365361,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 12454
+    },
+    {
+      "epoch": 0.6351234957926415,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 12455
+    },
+    {
+      "epoch": 0.6351744892487469,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 12456
+    },
+    {
+      "epoch": 0.6352254827048522,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12457
+    },
+    {
+      "epoch": 0.6352764761609576,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12458
+    },
+    {
+      "epoch": 0.6353274696170631,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 12459
+    },
+    {
+      "epoch": 0.6353784630731685,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6354294565292739,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 12461
+    },
+    {
+      "epoch": 0.6354804499853792,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 12462
+    },
+    {
+      "epoch": 0.6355314434414846,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 12463
+    },
+    {
+      "epoch": 0.63558243689759,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 12464
+    },
+    {
+      "epoch": 0.6356334303536954,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 12465
+    },
+    {
+      "epoch": 0.6356844238098008,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 12466
+    },
+    {
+      "epoch": 0.6357354172659062,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7802,
+      "step": 12467
+    },
+    {
+      "epoch": 0.6357864107220116,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 12468
+    },
+    {
+      "epoch": 0.635837404178117,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 12469
+    },
+    {
+      "epoch": 0.6358883976342223,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7726,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6359393910903277,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 12471
+    },
+    {
+      "epoch": 0.6359903845464331,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 12472
+    },
+    {
+      "epoch": 0.6360413780025386,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.771,
+      "step": 12473
+    },
+    {
+      "epoch": 0.636092371458644,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 12474
+    },
+    {
+      "epoch": 0.6361433649147493,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 12475
+    },
+    {
+      "epoch": 0.6361943583708547,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 12476
+    },
+    {
+      "epoch": 0.6362453518269601,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 12477
+    },
+    {
+      "epoch": 0.6362963452830654,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 12478
+    },
+    {
+      "epoch": 0.6363473387391708,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 12479
+    },
+    {
+      "epoch": 0.6363983321952763,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6364493256513817,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 12481
+    },
+    {
+      "epoch": 0.6365003191074871,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 12482
+    },
+    {
+      "epoch": 0.6365513125635924,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 12483
+    },
+    {
+      "epoch": 0.6366023060196978,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 12484
+    },
+    {
+      "epoch": 0.6366532994758032,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7727,
+      "step": 12485
+    },
+    {
+      "epoch": 0.6367042929319086,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 12486
+    },
+    {
+      "epoch": 0.636755286388014,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 12487
+    },
+    {
+      "epoch": 0.6368062798441194,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7798,
+      "step": 12488
+    },
+    {
+      "epoch": 0.6368572733002248,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 12489
+    },
+    {
+      "epoch": 0.6369082667563302,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6369592602124355,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 12491
+    },
+    {
+      "epoch": 0.6370102536685409,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 12492
+    },
+    {
+      "epoch": 0.6370612471246463,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 12493
+    },
+    {
+      "epoch": 0.6371122405807518,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 12494
+    },
+    {
+      "epoch": 0.6371632340368572,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 12495
+    },
+    {
+      "epoch": 0.6372142274929625,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 12496
+    },
+    {
+      "epoch": 0.6372652209490679,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 12497
+    },
+    {
+      "epoch": 0.6373162144051733,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 12498
+    },
+    {
+      "epoch": 0.6373672078612787,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 12499
+    },
+    {
+      "epoch": 0.637418201317384,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6374691947734895,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 12501
+    },
+    {
+      "epoch": 0.6375201882295949,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 12502
+    },
+    {
+      "epoch": 0.6375711816857003,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 12503
+    },
+    {
+      "epoch": 0.6376221751418056,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 12504
+    },
+    {
+      "epoch": 0.637673168597911,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 12505
+    },
+    {
+      "epoch": 0.6377241620540164,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 12506
+    },
+    {
+      "epoch": 0.6377751555101218,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 12507
+    },
+    {
+      "epoch": 0.6378261489662272,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7699,
+      "step": 12508
+    },
+    {
+      "epoch": 0.6378771424223326,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 12509
+    },
+    {
+      "epoch": 0.637928135878438,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7679,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6379791293345434,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 12511
+    },
+    {
+      "epoch": 0.6380301227906487,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 12512
+    },
+    {
+      "epoch": 0.6380811162467541,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 12513
+    },
+    {
+      "epoch": 0.6381321097028595,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 12514
+    },
+    {
+      "epoch": 0.638183103158965,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 12515
+    },
+    {
+      "epoch": 0.6382340966150704,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12516
+    },
+    {
+      "epoch": 0.6382850900711757,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 12517
+    },
+    {
+      "epoch": 0.6383360835272811,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 12518
+    },
+    {
+      "epoch": 0.6383870769833865,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 12519
+    },
+    {
+      "epoch": 0.6384380704394919,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6384890638955972,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 12521
+    },
+    {
+      "epoch": 0.6385400573517027,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 12522
+    },
+    {
+      "epoch": 0.6385910508078081,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12523
+    },
+    {
+      "epoch": 0.6386420442639135,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 12524
+    },
+    {
+      "epoch": 0.6386930377200188,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 12525
+    },
+    {
+      "epoch": 0.6387440311761242,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 12526
+    },
+    {
+      "epoch": 0.6387950246322296,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7854,
+      "step": 12527
+    },
+    {
+      "epoch": 0.638846018088335,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 12528
+    },
+    {
+      "epoch": 0.6388970115444405,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 12529
+    },
+    {
+      "epoch": 0.6389480050005458,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6389989984566512,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 12531
+    },
+    {
+      "epoch": 0.6390499919127566,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 12532
+    },
+    {
+      "epoch": 0.639100985368862,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 12533
+    },
+    {
+      "epoch": 0.6391519788249673,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 12534
+    },
+    {
+      "epoch": 0.6392029722810727,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 12535
+    },
+    {
+      "epoch": 0.6392539657371782,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 12536
+    },
+    {
+      "epoch": 0.6393049591932836,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 12537
+    },
+    {
+      "epoch": 0.6393559526493889,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 12538
+    },
+    {
+      "epoch": 0.6394069461054943,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 12539
+    },
+    {
+      "epoch": 0.6394579395615997,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7717,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6395089330177051,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7639,
+      "step": 12541
+    },
+    {
+      "epoch": 0.6395599264738104,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 12542
+    },
+    {
+      "epoch": 0.6396109199299159,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 12543
+    },
+    {
+      "epoch": 0.6396619133860213,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 12544
+    },
+    {
+      "epoch": 0.6397129068421267,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 12545
+    },
+    {
+      "epoch": 0.639763900298232,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 12546
+    },
+    {
+      "epoch": 0.6398148937543374,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 12547
+    },
+    {
+      "epoch": 0.6398658872104428,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 12548
+    },
+    {
+      "epoch": 0.6399168806665482,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 12549
+    },
+    {
+      "epoch": 0.6399678741226537,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 12550
+    },
+    {
+      "epoch": 0.640018867578759,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 12551
+    },
+    {
+      "epoch": 0.6400698610348644,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 12552
+    },
+    {
+      "epoch": 0.6401208544909698,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7737,
+      "step": 12553
+    },
+    {
+      "epoch": 0.6401718479470752,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 12554
+    },
+    {
+      "epoch": 0.6402228414031805,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 12555
+    },
+    {
+      "epoch": 0.6402738348592859,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.777,
+      "step": 12556
+    },
+    {
+      "epoch": 0.6403248283153914,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 12557
+    },
+    {
+      "epoch": 0.6403758217714968,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 12558
+    },
+    {
+      "epoch": 0.6404268152276021,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12559
+    },
+    {
+      "epoch": 0.6404778086837075,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7809,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6405288021398129,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8329,
+      "step": 12561
+    },
+    {
+      "epoch": 0.6405797955959183,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 12562
+    },
+    {
+      "epoch": 0.6406307890520236,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 12563
+    },
+    {
+      "epoch": 0.6406817825081291,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 12564
+    },
+    {
+      "epoch": 0.6407327759642345,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 12565
+    },
+    {
+      "epoch": 0.6407837694203399,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 12566
+    },
+    {
+      "epoch": 0.6408347628764453,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 12567
+    },
+    {
+      "epoch": 0.6408857563325506,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 12568
+    },
+    {
+      "epoch": 0.640936749788656,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 12569
+    },
+    {
+      "epoch": 0.6409877432447614,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6410387367008669,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12571
+    },
+    {
+      "epoch": 0.6410897301569722,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 12572
+    },
+    {
+      "epoch": 0.6411407236130776,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8428,
+      "step": 12573
+    },
+    {
+      "epoch": 0.641191717069183,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 12574
+    },
+    {
+      "epoch": 0.6412427105252884,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 12575
+    },
+    {
+      "epoch": 0.6412937039813937,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 12576
+    },
+    {
+      "epoch": 0.6413446974374991,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 12577
+    },
+    {
+      "epoch": 0.6413956908936046,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 12578
+    },
+    {
+      "epoch": 0.64144668434971,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 12579
+    },
+    {
+      "epoch": 0.6414976778058153,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6415486712619207,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 12581
+    },
+    {
+      "epoch": 0.6415996647180261,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 12582
+    },
+    {
+      "epoch": 0.6416506581741315,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 12583
+    },
+    {
+      "epoch": 0.6417016516302368,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 12584
+    },
+    {
+      "epoch": 0.6417526450863423,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 12585
+    },
+    {
+      "epoch": 0.6418036385424477,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 12586
+    },
+    {
+      "epoch": 0.6418546319985531,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 12587
+    },
+    {
+      "epoch": 0.6419056254546585,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 12588
+    },
+    {
+      "epoch": 0.6419566189107638,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 12589
+    },
+    {
+      "epoch": 0.6420076123668692,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.77,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6420586058229746,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 12591
+    },
+    {
+      "epoch": 0.6421095992790801,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7717,
+      "step": 12592
+    },
+    {
+      "epoch": 0.6421605927351854,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8264,
+      "step": 12593
+    },
+    {
+      "epoch": 0.6422115861912908,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 12594
+    },
+    {
+      "epoch": 0.6422625796473962,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 12595
+    },
+    {
+      "epoch": 0.6423135731035016,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 12596
+    },
+    {
+      "epoch": 0.6423645665596069,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 12597
+    },
+    {
+      "epoch": 0.6424155600157123,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 12598
+    },
+    {
+      "epoch": 0.6424665534718178,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 12599
+    },
+    {
+      "epoch": 0.6425175469279232,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6425685403840286,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12601
+    },
+    {
+      "epoch": 0.6426195338401339,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 12602
+    },
+    {
+      "epoch": 0.6426705272962393,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 12603
+    },
+    {
+      "epoch": 0.6427215207523447,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 12604
+    },
+    {
+      "epoch": 0.64277251420845,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 12605
+    },
+    {
+      "epoch": 0.6428235076645555,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 12606
+    },
+    {
+      "epoch": 0.6428745011206609,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 12607
+    },
+    {
+      "epoch": 0.6429254945767663,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 12608
+    },
+    {
+      "epoch": 0.6429764880328717,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 12609
+    },
+    {
+      "epoch": 0.643027481488977,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6430784749450824,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 12611
+    },
+    {
+      "epoch": 0.6431294684011878,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 12612
+    },
+    {
+      "epoch": 0.6431804618572933,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.77,
+      "step": 12613
+    },
+    {
+      "epoch": 0.6432314553133986,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 12614
+    },
+    {
+      "epoch": 0.643282448769504,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7739,
+      "step": 12615
+    },
+    {
+      "epoch": 0.6433334422256094,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 12616
+    },
+    {
+      "epoch": 0.6433844356817148,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7752,
+      "step": 12617
+    },
+    {
+      "epoch": 0.6434354291378201,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7637,
+      "step": 12618
+    },
+    {
+      "epoch": 0.6434864225939255,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 12619
+    },
+    {
+      "epoch": 0.643537416050031,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6435884095061364,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 12621
+    },
+    {
+      "epoch": 0.6436394029622418,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 12622
+    },
+    {
+      "epoch": 0.6436903964183471,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.779,
+      "step": 12623
+    },
+    {
+      "epoch": 0.6437413898744525,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 12624
+    },
+    {
+      "epoch": 0.6437923833305579,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12625
+    },
+    {
+      "epoch": 0.6438433767866633,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 12626
+    },
+    {
+      "epoch": 0.6438943702427687,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 12627
+    },
+    {
+      "epoch": 0.6439453636988741,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7818,
+      "step": 12628
+    },
+    {
+      "epoch": 0.6439963571549795,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 12629
+    },
+    {
+      "epoch": 0.6440473506110849,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6440983440671902,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7765,
+      "step": 12631
+    },
+    {
+      "epoch": 0.6441493375232956,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 12632
+    },
+    {
+      "epoch": 0.644200330979401,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 12633
+    },
+    {
+      "epoch": 0.6442513244355065,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 12634
+    },
+    {
+      "epoch": 0.6443023178916119,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 12635
+    },
+    {
+      "epoch": 0.6443533113477172,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 12636
+    },
+    {
+      "epoch": 0.6444043048038226,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 12637
+    },
+    {
+      "epoch": 0.644455298259928,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 12638
+    },
+    {
+      "epoch": 0.6445062917160334,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 12639
+    },
+    {
+      "epoch": 0.6445572851721387,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6446082786282442,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 12641
+    },
+    {
+      "epoch": 0.6446592720843496,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 12642
+    },
+    {
+      "epoch": 0.644710265540455,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 12643
+    },
+    {
+      "epoch": 0.6447612589965603,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7792,
+      "step": 12644
+    },
+    {
+      "epoch": 0.6448122524526657,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 12645
+    },
+    {
+      "epoch": 0.6448632459087711,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 12646
+    },
+    {
+      "epoch": 0.6449142393648765,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7721,
+      "step": 12647
+    },
+    {
+      "epoch": 0.644965232820982,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12648
+    },
+    {
+      "epoch": 0.6450162262770873,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 12649
+    },
+    {
+      "epoch": 0.6450672197331927,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6451182131892981,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7766,
+      "step": 12651
+    },
+    {
+      "epoch": 0.6451692066454034,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 12652
+    },
+    {
+      "epoch": 0.6452202001015088,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 12653
+    },
+    {
+      "epoch": 0.6452711935576142,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 12654
+    },
+    {
+      "epoch": 0.6453221870137197,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 12655
+    },
+    {
+      "epoch": 0.6453731804698251,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 12656
+    },
+    {
+      "epoch": 0.6454241739259304,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 12657
+    },
+    {
+      "epoch": 0.6454751673820358,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 12658
+    },
+    {
+      "epoch": 0.6455261608381412,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 12659
+    },
+    {
+      "epoch": 0.6455771542942466,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6456281477503519,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 12661
+    },
+    {
+      "epoch": 0.6456791412064574,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 12662
+    },
+    {
+      "epoch": 0.6457301346625628,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 12663
+    },
+    {
+      "epoch": 0.6457811281186682,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7702,
+      "step": 12664
+    },
+    {
+      "epoch": 0.6458321215747735,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 12665
+    },
+    {
+      "epoch": 0.6458831150308789,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 12666
+    },
+    {
+      "epoch": 0.6459341084869843,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 12667
+    },
+    {
+      "epoch": 0.6459851019430897,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 12668
+    },
+    {
+      "epoch": 0.6460360953991952,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 12669
+    },
+    {
+      "epoch": 0.6460870888553005,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6461380823114059,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 12671
+    },
+    {
+      "epoch": 0.6461890757675113,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 12672
+    },
+    {
+      "epoch": 0.6462400692236167,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 12673
+    },
+    {
+      "epoch": 0.646291062679722,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 12674
+    },
+    {
+      "epoch": 0.6463420561358274,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 12675
+    },
+    {
+      "epoch": 0.6463930495919329,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 12676
+    },
+    {
+      "epoch": 0.6464440430480383,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 12677
+    },
+    {
+      "epoch": 0.6464950365041436,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7801,
+      "step": 12678
+    },
+    {
+      "epoch": 0.646546029960249,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 12679
+    },
+    {
+      "epoch": 0.6465970234163544,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6466480168724598,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 12681
+    },
+    {
+      "epoch": 0.6466990103285651,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 12682
+    },
+    {
+      "epoch": 0.6467500037846706,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 12683
+    },
+    {
+      "epoch": 0.646800997240776,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 12684
+    },
+    {
+      "epoch": 0.6468519906968814,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 12685
+    },
+    {
+      "epoch": 0.6469029841529867,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 12686
+    },
+    {
+      "epoch": 0.6469539776090921,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 12687
+    },
+    {
+      "epoch": 0.6470049710651975,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 12688
+    },
+    {
+      "epoch": 0.6470559645213029,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 12689
+    },
+    {
+      "epoch": 0.6471069579774084,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6471579514335137,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 12691
+    },
+    {
+      "epoch": 0.6472089448896191,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7671,
+      "step": 12692
+    },
+    {
+      "epoch": 0.6472599383457245,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 12693
+    },
+    {
+      "epoch": 0.6473109318018299,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12694
+    },
+    {
+      "epoch": 0.6473619252579352,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7722,
+      "step": 12695
+    },
+    {
+      "epoch": 0.6474129187140406,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7764,
+      "step": 12696
+    },
+    {
+      "epoch": 0.6474639121701461,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 12697
+    },
+    {
+      "epoch": 0.6475149056262515,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 12698
+    },
+    {
+      "epoch": 0.6475658990823568,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7714,
+      "step": 12699
+    },
+    {
+      "epoch": 0.6476168925384622,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6476678859945676,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 12701
+    },
+    {
+      "epoch": 0.647718879450673,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7893,
+      "step": 12702
+    },
+    {
+      "epoch": 0.6477698729067783,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 12703
+    },
+    {
+      "epoch": 0.6478208663628838,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 12704
+    },
+    {
+      "epoch": 0.6478718598189892,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 12705
+    },
+    {
+      "epoch": 0.6479228532750946,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 12706
+    },
+    {
+      "epoch": 0.6479738467312,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 12707
+    },
+    {
+      "epoch": 0.6480248401873053,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 12708
+    },
+    {
+      "epoch": 0.6480758336434107,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 12709
+    },
+    {
+      "epoch": 0.6481268270995161,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6481778205556216,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 12711
+    },
+    {
+      "epoch": 0.6482288140117269,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 12712
+    },
+    {
+      "epoch": 0.6482798074678323,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 12713
+    },
+    {
+      "epoch": 0.6483308009239377,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 12714
+    },
+    {
+      "epoch": 0.6483817943800431,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 12715
+    },
+    {
+      "epoch": 0.6484327878361484,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7808,
+      "step": 12716
+    },
+    {
+      "epoch": 0.6484837812922538,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 12717
+    },
+    {
+      "epoch": 0.6485347747483593,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7719,
+      "step": 12718
+    },
+    {
+      "epoch": 0.6485857682044647,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 12719
+    },
+    {
+      "epoch": 0.64863676166057,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6486877551166754,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 12721
+    },
+    {
+      "epoch": 0.6487387485727808,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 12722
+    },
+    {
+      "epoch": 0.6487897420288862,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 12723
+    },
+    {
+      "epoch": 0.6488407354849915,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 12724
+    },
+    {
+      "epoch": 0.648891728941097,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 12725
+    },
+    {
+      "epoch": 0.6489427223972024,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 12726
+    },
+    {
+      "epoch": 0.6489937158533078,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 12727
+    },
+    {
+      "epoch": 0.6490447093094132,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 12728
+    },
+    {
+      "epoch": 0.6490957027655185,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 12729
+    },
+    {
+      "epoch": 0.6491466962216239,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6491976896777293,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 12731
+    },
+    {
+      "epoch": 0.6492486831338348,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 12732
+    },
+    {
+      "epoch": 0.6492996765899401,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 12733
+    },
+    {
+      "epoch": 0.6493506700460455,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 12734
+    },
+    {
+      "epoch": 0.6494016635021509,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7503,
+      "step": 12735
+    },
+    {
+      "epoch": 0.6494526569582563,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 12736
+    },
+    {
+      "epoch": 0.6495036504143616,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 12737
+    },
+    {
+      "epoch": 0.649554643870467,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 12738
+    },
+    {
+      "epoch": 0.6496056373265725,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 12739
+    },
+    {
+      "epoch": 0.6496566307826779,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6497076242387833,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12741
+    },
+    {
+      "epoch": 0.6497586176948886,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 12742
+    },
+    {
+      "epoch": 0.649809611150994,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 12743
+    },
+    {
+      "epoch": 0.6498606046070994,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 12744
+    },
+    {
+      "epoch": 0.6499115980632048,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 12745
+    },
+    {
+      "epoch": 0.6499625915193102,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 12746
+    },
+    {
+      "epoch": 0.6500135849754156,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 12747
+    },
+    {
+      "epoch": 0.650064578431521,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 12748
+    },
+    {
+      "epoch": 0.6501155718876264,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7741,
+      "step": 12749
+    },
+    {
+      "epoch": 0.6501665653437317,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6502175587998371,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 12751
+    },
+    {
+      "epoch": 0.6502685522559425,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 12752
+    },
+    {
+      "epoch": 0.650319545712048,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 12753
+    },
+    {
+      "epoch": 0.6503705391681533,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7805,
+      "step": 12754
+    },
+    {
+      "epoch": 0.6504215326242587,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 12755
+    },
+    {
+      "epoch": 0.6504725260803641,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 12756
+    },
+    {
+      "epoch": 0.6505235195364695,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 12757
+    },
+    {
+      "epoch": 0.6505745129925748,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 12758
+    },
+    {
+      "epoch": 0.6506255064486802,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 12759
+    },
+    {
+      "epoch": 0.6506764999047857,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6507274933608911,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 12761
+    },
+    {
+      "epoch": 0.6507784868169965,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 12762
+    },
+    {
+      "epoch": 0.6508294802731018,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 12763
+    },
+    {
+      "epoch": 0.6508804737292072,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 12764
+    },
+    {
+      "epoch": 0.6509314671853126,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 12765
+    },
+    {
+      "epoch": 0.650982460641418,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 12766
+    },
+    {
+      "epoch": 0.6510334540975234,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 12767
+    },
+    {
+      "epoch": 0.6510844475536288,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 12768
+    },
+    {
+      "epoch": 0.6511354410097342,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 12769
+    },
+    {
+      "epoch": 0.6511864344658396,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 12770
+    },
+    {
+      "epoch": 0.651237427921945,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 12771
+    },
+    {
+      "epoch": 0.6512884213780503,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 12772
+    },
+    {
+      "epoch": 0.6513394148341557,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7746,
+      "step": 12773
+    },
+    {
+      "epoch": 0.6513904082902612,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 12774
+    },
+    {
+      "epoch": 0.6514414017463666,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 12775
+    },
+    {
+      "epoch": 0.6514923952024719,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 12776
+    },
+    {
+      "epoch": 0.6515433886585773,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 12777
+    },
+    {
+      "epoch": 0.6515943821146827,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 12778
+    },
+    {
+      "epoch": 0.651645375570788,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 12779
+    },
+    {
+      "epoch": 0.6516963690268934,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 12780
+    },
+    {
+      "epoch": 0.6517473624829989,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 12781
+    },
+    {
+      "epoch": 0.6517983559391043,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 12782
+    },
+    {
+      "epoch": 0.6518493493952097,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7683,
+      "step": 12783
+    },
+    {
+      "epoch": 0.651900342851315,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 12784
+    },
+    {
+      "epoch": 0.6519513363074204,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 12785
+    },
+    {
+      "epoch": 0.6520023297635258,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 12786
+    },
+    {
+      "epoch": 0.6520533232196312,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 12787
+    },
+    {
+      "epoch": 0.6521043166757367,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 12788
+    },
+    {
+      "epoch": 0.652155310131842,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 12789
+    },
+    {
+      "epoch": 0.6522063035879474,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6522572970440528,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 12791
+    },
+    {
+      "epoch": 0.6523082905001581,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 12792
+    },
+    {
+      "epoch": 0.6523592839562635,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 12793
+    },
+    {
+      "epoch": 0.6524102774123689,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 12794
+    },
+    {
+      "epoch": 0.6524612708684744,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 12795
+    },
+    {
+      "epoch": 0.6525122643245798,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12796
+    },
+    {
+      "epoch": 0.6525632577806851,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 12797
+    },
+    {
+      "epoch": 0.6526142512367905,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 12798
+    },
+    {
+      "epoch": 0.6526652446928959,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7752,
+      "step": 12799
+    },
+    {
+      "epoch": 0.6527162381490013,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6527672316051066,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 12801
+    },
+    {
+      "epoch": 0.6528182250612121,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 12802
+    },
+    {
+      "epoch": 0.6528692185173175,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 12803
+    },
+    {
+      "epoch": 0.6529202119734229,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 12804
+    },
+    {
+      "epoch": 0.6529712054295282,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 12805
+    },
+    {
+      "epoch": 0.6530221988856336,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 12806
+    },
+    {
+      "epoch": 0.653073192341739,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7759,
+      "step": 12807
+    },
+    {
+      "epoch": 0.6531241857978444,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 12808
+    },
+    {
+      "epoch": 0.6531751792539499,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 12809
+    },
+    {
+      "epoch": 0.6532261727100552,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6532771661661606,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7734,
+      "step": 12811
+    },
+    {
+      "epoch": 0.653328159622266,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 12812
+    },
+    {
+      "epoch": 0.6533791530783714,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 12813
+    },
+    {
+      "epoch": 0.6534301465344767,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7724,
+      "step": 12814
+    },
+    {
+      "epoch": 0.6534811399905821,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 12815
+    },
+    {
+      "epoch": 0.6535321334466876,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 12816
+    },
+    {
+      "epoch": 0.653583126902793,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 12817
+    },
+    {
+      "epoch": 0.6536341203588983,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 12818
+    },
+    {
+      "epoch": 0.6536851138150037,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 12819
+    },
+    {
+      "epoch": 0.6537361072711091,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6537871007272145,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 12821
+    },
+    {
+      "epoch": 0.6538380941833198,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 12822
+    },
+    {
+      "epoch": 0.6538890876394253,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 12823
+    },
+    {
+      "epoch": 0.6539400810955307,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 12824
+    },
+    {
+      "epoch": 0.6539910745516361,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 12825
+    },
+    {
+      "epoch": 0.6540420680077415,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 12826
+    },
+    {
+      "epoch": 0.6540930614638468,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 12827
+    },
+    {
+      "epoch": 0.6541440549199522,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 12828
+    },
+    {
+      "epoch": 0.6541950483760576,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 12829
+    },
+    {
+      "epoch": 0.6542460418321631,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6542970352882684,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7769,
+      "step": 12831
+    },
+    {
+      "epoch": 0.6543480287443738,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 12832
+    },
+    {
+      "epoch": 0.6543990222004792,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 12833
+    },
+    {
+      "epoch": 0.6544500156565846,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.762,
+      "step": 12834
+    },
+    {
+      "epoch": 0.6545010091126899,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 12835
+    },
+    {
+      "epoch": 0.6545520025687953,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 12836
+    },
+    {
+      "epoch": 0.6546029960249008,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 12837
+    },
+    {
+      "epoch": 0.6546539894810062,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 12838
+    },
+    {
+      "epoch": 0.6547049829371115,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 12839
+    },
+    {
+      "epoch": 0.6547559763932169,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6548069698493223,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 12841
+    },
+    {
+      "epoch": 0.6548579633054277,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 12842
+    },
+    {
+      "epoch": 0.654908956761533,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8151,
+      "step": 12843
+    },
+    {
+      "epoch": 0.6549599502176385,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 12844
+    },
+    {
+      "epoch": 0.6550109436737439,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 12845
+    },
+    {
+      "epoch": 0.6550619371298493,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 12846
+    },
+    {
+      "epoch": 0.6551129305859547,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 12847
+    },
+    {
+      "epoch": 0.65516392404206,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 12848
+    },
+    {
+      "epoch": 0.6552149174981654,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7872,
+      "step": 12849
+    },
+    {
+      "epoch": 0.6552659109542708,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6553169044103763,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7675,
+      "step": 12851
+    },
+    {
+      "epoch": 0.6553678978664816,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 12852
+    },
+    {
+      "epoch": 0.655418891322587,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 12853
+    },
+    {
+      "epoch": 0.6554698847786924,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 12854
+    },
+    {
+      "epoch": 0.6555208782347978,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7786,
+      "step": 12855
+    },
+    {
+      "epoch": 0.6555718716909031,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 12856
+    },
+    {
+      "epoch": 0.6556228651470085,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 12857
+    },
+    {
+      "epoch": 0.655673858603114,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 12858
+    },
+    {
+      "epoch": 0.6557248520592194,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 12859
+    },
+    {
+      "epoch": 0.6557758455153248,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6558268389714301,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 12861
+    },
+    {
+      "epoch": 0.6558778324275355,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7765,
+      "step": 12862
+    },
+    {
+      "epoch": 0.6559288258836409,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 12863
+    },
+    {
+      "epoch": 0.6559798193397463,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 12864
+    },
+    {
+      "epoch": 0.6560308127958517,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7878,
+      "step": 12865
+    },
+    {
+      "epoch": 0.6560818062519571,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 12866
+    },
+    {
+      "epoch": 0.6561327997080625,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 12867
+    },
+    {
+      "epoch": 0.6561837931641679,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 12868
+    },
+    {
+      "epoch": 0.6562347866202732,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 12869
+    },
+    {
+      "epoch": 0.6562857800763786,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 12870
+    },
+    {
+      "epoch": 0.656336773532484,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 12871
+    },
+    {
+      "epoch": 0.6563877669885895,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 12872
+    },
+    {
+      "epoch": 0.6564387604446948,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 12873
+    },
+    {
+      "epoch": 0.6564897539008002,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7747,
+      "step": 12874
+    },
+    {
+      "epoch": 0.6565407473569056,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 12875
+    },
+    {
+      "epoch": 0.656591740813011,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 12876
+    },
+    {
+      "epoch": 0.6566427342691163,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12877
+    },
+    {
+      "epoch": 0.6566937277252217,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 12878
+    },
+    {
+      "epoch": 0.6567447211813272,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 12879
+    },
+    {
+      "epoch": 0.6567957146374326,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 12880
+    },
+    {
+      "epoch": 0.656846708093538,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 12881
+    },
+    {
+      "epoch": 0.6568977015496433,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 12882
+    },
+    {
+      "epoch": 0.6569486950057487,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7858,
+      "step": 12883
+    },
+    {
+      "epoch": 0.6569996884618541,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.774,
+      "step": 12884
+    },
+    {
+      "epoch": 0.6570506819179595,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 12885
+    },
+    {
+      "epoch": 0.6571016753740649,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7903,
+      "step": 12886
+    },
+    {
+      "epoch": 0.6571526688301703,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 12887
+    },
+    {
+      "epoch": 0.6572036622862757,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7758,
+      "step": 12888
+    },
+    {
+      "epoch": 0.6572546557423811,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 12889
+    },
+    {
+      "epoch": 0.6573056491984864,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6573566426545918,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 12891
+    },
+    {
+      "epoch": 0.6574076361106972,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 12892
+    },
+    {
+      "epoch": 0.6574586295668027,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 12893
+    },
+    {
+      "epoch": 0.657509623022908,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 12894
+    },
+    {
+      "epoch": 0.6575606164790134,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 12895
+    },
+    {
+      "epoch": 0.6576116099351188,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7738,
+      "step": 12896
+    },
+    {
+      "epoch": 0.6576626033912242,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7631,
+      "step": 12897
+    },
+    {
+      "epoch": 0.6577135968473296,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 12898
+    },
+    {
+      "epoch": 0.6577645903034349,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 12899
+    },
+    {
+      "epoch": 0.6578155837595404,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6578665772156458,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 12901
+    },
+    {
+      "epoch": 0.6579175706717512,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7782,
+      "step": 12902
+    },
+    {
+      "epoch": 0.6579685641278565,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 12903
+    },
+    {
+      "epoch": 0.6580195575839619,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 12904
+    },
+    {
+      "epoch": 0.6580705510400673,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 12905
+    },
+    {
+      "epoch": 0.6581215444961727,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 12906
+    },
+    {
+      "epoch": 0.6581725379522781,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 12907
+    },
+    {
+      "epoch": 0.6582235314083835,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 12908
+    },
+    {
+      "epoch": 0.6582745248644889,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7809,
+      "step": 12909
+    },
+    {
+      "epoch": 0.6583255183205943,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6583765117766996,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 12911
+    },
+    {
+      "epoch": 0.658427505232805,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 12912
+    },
+    {
+      "epoch": 0.6584784986889104,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 12913
+    },
+    {
+      "epoch": 0.6585294921450159,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 12914
+    },
+    {
+      "epoch": 0.6585804856011213,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 12915
+    },
+    {
+      "epoch": 0.6586314790572266,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 12916
+    },
+    {
+      "epoch": 0.658682472513332,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 12917
+    },
+    {
+      "epoch": 0.6587334659694374,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 12918
+    },
+    {
+      "epoch": 0.6587844594255428,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 12919
+    },
+    {
+      "epoch": 0.6588354528816481,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6588864463377536,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 12921
+    },
+    {
+      "epoch": 0.658937439793859,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 12922
+    },
+    {
+      "epoch": 0.6589884332499644,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7595,
+      "step": 12923
+    },
+    {
+      "epoch": 0.6590394267060697,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 12924
+    },
+    {
+      "epoch": 0.6590904201621751,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 12925
+    },
+    {
+      "epoch": 0.6591414136182805,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 12926
+    },
+    {
+      "epoch": 0.6591924070743859,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 12927
+    },
+    {
+      "epoch": 0.6592434005304914,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 12928
+    },
+    {
+      "epoch": 0.6592943939865967,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7812,
+      "step": 12929
+    },
+    {
+      "epoch": 0.6593453874427021,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6593963808988075,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7463,
+      "step": 12931
+    },
+    {
+      "epoch": 0.6594473743549129,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 12932
+    },
+    {
+      "epoch": 0.6594983678110182,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 12933
+    },
+    {
+      "epoch": 0.6595493612671236,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 12934
+    },
+    {
+      "epoch": 0.6596003547232291,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 12935
+    },
+    {
+      "epoch": 0.6596513481793345,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7688,
+      "step": 12936
+    },
+    {
+      "epoch": 0.6597023416354398,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 12937
+    },
+    {
+      "epoch": 0.6597533350915452,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 12938
+    },
+    {
+      "epoch": 0.6598043285476506,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 12939
+    },
+    {
+      "epoch": 0.659855322003756,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6599063154598613,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 12941
+    },
+    {
+      "epoch": 0.6599573089159668,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 12942
+    },
+    {
+      "epoch": 0.6600083023720722,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 12943
+    },
+    {
+      "epoch": 0.6600592958281776,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 12944
+    },
+    {
+      "epoch": 0.660110289284283,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 12945
+    },
+    {
+      "epoch": 0.6601612827403883,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 12946
+    },
+    {
+      "epoch": 0.6602122761964937,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 12947
+    },
+    {
+      "epoch": 0.6602632696525991,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 12948
+    },
+    {
+      "epoch": 0.6603142631087046,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 12949
+    },
+    {
+      "epoch": 0.6603652565648099,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6604162500209153,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7866,
+      "step": 12951
+    },
+    {
+      "epoch": 0.6604672434770207,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7928,
+      "step": 12952
+    },
+    {
+      "epoch": 0.6605182369331261,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 12953
+    },
+    {
+      "epoch": 0.6605692303892314,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 12954
+    },
+    {
+      "epoch": 0.6606202238453368,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 12955
+    },
+    {
+      "epoch": 0.6606712173014423,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 12956
+    },
+    {
+      "epoch": 0.6607222107575477,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 12957
+    },
+    {
+      "epoch": 0.660773204213653,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 12958
+    },
+    {
+      "epoch": 0.6608241976697584,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 12959
+    },
+    {
+      "epoch": 0.6608751911258638,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6609261845819692,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7622,
+      "step": 12961
+    },
+    {
+      "epoch": 0.6609771780380745,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 12962
+    },
+    {
+      "epoch": 0.66102817149418,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 12963
+    },
+    {
+      "epoch": 0.6610791649502854,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 12964
+    },
+    {
+      "epoch": 0.6611301584063908,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 12965
+    },
+    {
+      "epoch": 0.6611811518624962,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 12966
+    },
+    {
+      "epoch": 0.6612321453186015,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 12967
+    },
+    {
+      "epoch": 0.6612831387747069,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 12968
+    },
+    {
+      "epoch": 0.6613341322308123,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 12969
+    },
+    {
+      "epoch": 0.6613851256869178,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7732,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6614361191430231,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 12971
+    },
+    {
+      "epoch": 0.6614871125991285,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 12972
+    },
+    {
+      "epoch": 0.6615381060552339,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.802,
+      "step": 12973
+    },
+    {
+      "epoch": 0.6615890995113393,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 12974
+    },
+    {
+      "epoch": 0.6616400929674446,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 12975
+    },
+    {
+      "epoch": 0.66169108642355,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 12976
+    },
+    {
+      "epoch": 0.6617420798796555,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 12977
+    },
+    {
+      "epoch": 0.6617930733357609,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 12978
+    },
+    {
+      "epoch": 0.6618440667918662,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 12979
+    },
+    {
+      "epoch": 0.6618950602479716,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 12980
+    },
+    {
+      "epoch": 0.661946053704077,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 12981
+    },
+    {
+      "epoch": 0.6619970471601824,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 12982
+    },
+    {
+      "epoch": 0.6620480406162877,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7739,
+      "step": 12983
+    },
+    {
+      "epoch": 0.6620990340723932,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 12984
+    },
+    {
+      "epoch": 0.6621500275284986,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 12985
+    },
+    {
+      "epoch": 0.662201020984604,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 12986
+    },
+    {
+      "epoch": 0.6622520144407094,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 12987
+    },
+    {
+      "epoch": 0.6623030078968147,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 12988
+    },
+    {
+      "epoch": 0.6623540013529201,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7866,
+      "step": 12989
+    },
+    {
+      "epoch": 0.6624049948090255,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 12990
+    },
+    {
+      "epoch": 0.662455988265131,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 12991
+    },
+    {
+      "epoch": 0.6625069817212363,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 12992
+    },
+    {
+      "epoch": 0.6625579751773417,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 12993
+    },
+    {
+      "epoch": 0.6626089686334471,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 12994
+    },
+    {
+      "epoch": 0.6626599620895525,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 12995
+    },
+    {
+      "epoch": 0.6627109555456578,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 12996
+    },
+    {
+      "epoch": 0.6627619490017632,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 12997
+    },
+    {
+      "epoch": 0.6628129424578687,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 12998
+    },
+    {
+      "epoch": 0.6628639359139741,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 12999
+    },
+    {
+      "epoch": 0.6629149293700795,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6629659228261848,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7716,
+      "step": 13001
+    },
+    {
+      "epoch": 0.6630169162822902,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 13002
+    },
+    {
+      "epoch": 0.6630679097383956,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13003
+    },
+    {
+      "epoch": 0.663118903194501,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 13004
+    },
+    {
+      "epoch": 0.6631698966506064,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 13005
+    },
+    {
+      "epoch": 0.6632208901067118,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13006
+    },
+    {
+      "epoch": 0.6632718835628172,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 13007
+    },
+    {
+      "epoch": 0.6633228770189226,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 13008
+    },
+    {
+      "epoch": 0.6633738704750279,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 13009
+    },
+    {
+      "epoch": 0.6634248639311333,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6634758573872387,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 13011
+    },
+    {
+      "epoch": 0.6635268508433442,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 13012
+    },
+    {
+      "epoch": 0.6635778442994495,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 13013
+    },
+    {
+      "epoch": 0.6636288377555549,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 13014
+    },
+    {
+      "epoch": 0.6636798312116603,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 13015
+    },
+    {
+      "epoch": 0.6637308246677657,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 13016
+    },
+    {
+      "epoch": 0.663781818123871,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 13017
+    },
+    {
+      "epoch": 0.6638328115799764,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 13018
+    },
+    {
+      "epoch": 0.6638838050360819,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 13019
+    },
+    {
+      "epoch": 0.6639347984921873,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6639857919482927,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 13021
+    },
+    {
+      "epoch": 0.664036785404398,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 13022
+    },
+    {
+      "epoch": 0.6640877788605034,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.774,
+      "step": 13023
+    },
+    {
+      "epoch": 0.6641387723166088,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7807,
+      "step": 13024
+    },
+    {
+      "epoch": 0.6641897657727142,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 13025
+    },
+    {
+      "epoch": 0.6642407592288196,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7707,
+      "step": 13026
+    },
+    {
+      "epoch": 0.664291752684925,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8463,
+      "step": 13027
+    },
+    {
+      "epoch": 0.6643427461410304,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 13028
+    },
+    {
+      "epoch": 0.6643937395971358,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 13029
+    },
+    {
+      "epoch": 0.6644447330532411,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6644957265093465,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13031
+    },
+    {
+      "epoch": 0.6645467199654519,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 13032
+    },
+    {
+      "epoch": 0.6645977134215574,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 13033
+    },
+    {
+      "epoch": 0.6646487068776628,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 13034
+    },
+    {
+      "epoch": 0.6646997003337681,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 13035
+    },
+    {
+      "epoch": 0.6647506937898735,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 13036
+    },
+    {
+      "epoch": 0.6648016872459789,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 13037
+    },
+    {
+      "epoch": 0.6648526807020843,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 13038
+    },
+    {
+      "epoch": 0.6649036741581896,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 13039
+    },
+    {
+      "epoch": 0.6649546676142951,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6650056610704005,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 13041
+    },
+    {
+      "epoch": 0.6650566545265059,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13042
+    },
+    {
+      "epoch": 0.6651076479826112,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7771,
+      "step": 13043
+    },
+    {
+      "epoch": 0.6651586414387166,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 13044
+    },
+    {
+      "epoch": 0.665209634894822,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 13045
+    },
+    {
+      "epoch": 0.6652606283509274,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 13046
+    },
+    {
+      "epoch": 0.6653116218070328,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13047
+    },
+    {
+      "epoch": 0.6653626152631382,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 13048
+    },
+    {
+      "epoch": 0.6654136087192436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 13049
+    },
+    {
+      "epoch": 0.665464602175349,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7865,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6655155956314543,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 13051
+    },
+    {
+      "epoch": 0.6655665890875597,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 13052
+    },
+    {
+      "epoch": 0.6656175825436651,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7773,
+      "step": 13053
+    },
+    {
+      "epoch": 0.6656685759997706,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 13054
+    },
+    {
+      "epoch": 0.665719569455876,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 13055
+    },
+    {
+      "epoch": 0.6657705629119813,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 13056
+    },
+    {
+      "epoch": 0.6658215563680867,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 13057
+    },
+    {
+      "epoch": 0.6658725498241921,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7657,
+      "step": 13058
+    },
+    {
+      "epoch": 0.6659235432802975,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 13059
+    },
+    {
+      "epoch": 0.6659745367364028,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6660255301925083,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 13061
+    },
+    {
+      "epoch": 0.6660765236486137,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 13062
+    },
+    {
+      "epoch": 0.6661275171047191,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7862,
+      "step": 13063
+    },
+    {
+      "epoch": 0.6661785105608244,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 13064
+    },
+    {
+      "epoch": 0.6662295040169298,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 13065
+    },
+    {
+      "epoch": 0.6662804974730352,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 13066
+    },
+    {
+      "epoch": 0.6663314909291406,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 13067
+    },
+    {
+      "epoch": 0.666382484385246,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 13068
+    },
+    {
+      "epoch": 0.6664334778413514,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13069
+    },
+    {
+      "epoch": 0.6664844712974568,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.782,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6665354647535622,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 13071
+    },
+    {
+      "epoch": 0.6665864582096676,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 13072
+    },
+    {
+      "epoch": 0.6666374516657729,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 13073
+    },
+    {
+      "epoch": 0.6666884451218783,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 13074
+    },
+    {
+      "epoch": 0.6667394385779838,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 13075
+    },
+    {
+      "epoch": 0.6667904320340892,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 13076
+    },
+    {
+      "epoch": 0.6668414254901945,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13077
+    },
+    {
+      "epoch": 0.6668924189462999,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7712,
+      "step": 13078
+    },
+    {
+      "epoch": 0.6669434124024053,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 13079
+    },
+    {
+      "epoch": 0.6669944058585107,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 13080
+    },
+    {
+      "epoch": 0.667045399314616,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 13081
+    },
+    {
+      "epoch": 0.6670963927707215,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 13082
+    },
+    {
+      "epoch": 0.6671473862268269,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 13083
+    },
+    {
+      "epoch": 0.6671983796829323,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 13084
+    },
+    {
+      "epoch": 0.6672493731390376,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 13085
+    },
+    {
+      "epoch": 0.667300366595143,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 13086
+    },
+    {
+      "epoch": 0.6673513600512484,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13087
+    },
+    {
+      "epoch": 0.6674023535073538,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 13088
+    },
+    {
+      "epoch": 0.6674533469634593,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7794,
+      "step": 13089
+    },
+    {
+      "epoch": 0.6675043404195646,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 13090
+    },
+    {
+      "epoch": 0.66755533387567,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 13091
+    },
+    {
+      "epoch": 0.6676063273317754,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 13092
+    },
+    {
+      "epoch": 0.6676573207878808,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7797,
+      "step": 13093
+    },
+    {
+      "epoch": 0.6677083142439861,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 13094
+    },
+    {
+      "epoch": 0.6677593077000915,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13095
+    },
+    {
+      "epoch": 0.667810301156197,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 13096
+    },
+    {
+      "epoch": 0.6678612946123024,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 13097
+    },
+    {
+      "epoch": 0.6679122880684077,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 13098
+    },
+    {
+      "epoch": 0.6679632815245131,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 13099
+    },
+    {
+      "epoch": 0.6680142749806185,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6680652684367239,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 13101
+    },
+    {
+      "epoch": 0.6681162618928292,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 13102
+    },
+    {
+      "epoch": 0.6681672553489347,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 13103
+    },
+    {
+      "epoch": 0.6682182488050401,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 13104
+    },
+    {
+      "epoch": 0.6682692422611455,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 13105
+    },
+    {
+      "epoch": 0.6683202357172509,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 13106
+    },
+    {
+      "epoch": 0.6683712291733562,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 13107
+    },
+    {
+      "epoch": 0.6684222226294616,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13108
+    },
+    {
+      "epoch": 0.668473216085567,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 13109
+    },
+    {
+      "epoch": 0.6685242095416725,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6685752029977778,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 13111
+    },
+    {
+      "epoch": 0.6686261964538832,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7672,
+      "step": 13112
+    },
+    {
+      "epoch": 0.6686771899099886,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 13113
+    },
+    {
+      "epoch": 0.668728183366094,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7879,
+      "step": 13114
+    },
+    {
+      "epoch": 0.6687791768221993,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 13115
+    },
+    {
+      "epoch": 0.6688301702783047,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 13116
+    },
+    {
+      "epoch": 0.6688811637344102,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 13117
+    },
+    {
+      "epoch": 0.6689321571905156,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 13118
+    },
+    {
+      "epoch": 0.668983150646621,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13119
+    },
+    {
+      "epoch": 0.6690341441027263,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6690851375588317,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 13121
+    },
+    {
+      "epoch": 0.6691361310149371,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 13122
+    },
+    {
+      "epoch": 0.6691871244710424,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 13123
+    },
+    {
+      "epoch": 0.6692381179271479,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7814,
+      "step": 13124
+    },
+    {
+      "epoch": 0.6692891113832533,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 13125
+    },
+    {
+      "epoch": 0.6693401048393587,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7611,
+      "step": 13126
+    },
+    {
+      "epoch": 0.6693910982954641,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 13127
+    },
+    {
+      "epoch": 0.6694420917515694,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 13128
+    },
+    {
+      "epoch": 0.6694930852076748,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 13129
+    },
+    {
+      "epoch": 0.6695440786637802,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.782,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6695950721198857,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 13131
+    },
+    {
+      "epoch": 0.669646065575991,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 13132
+    },
+    {
+      "epoch": 0.6696970590320964,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7697,
+      "step": 13133
+    },
+    {
+      "epoch": 0.6697480524882018,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 13134
+    },
+    {
+      "epoch": 0.6697990459443072,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 13135
+    },
+    {
+      "epoch": 0.6698500394004125,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 13136
+    },
+    {
+      "epoch": 0.6699010328565179,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 13137
+    },
+    {
+      "epoch": 0.6699520263126234,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 13138
+    },
+    {
+      "epoch": 0.6700030197687288,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 13139
+    },
+    {
+      "epoch": 0.6700540132248342,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6701050066809395,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8323,
+      "step": 13141
+    },
+    {
+      "epoch": 0.6701560001370449,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 13142
+    },
+    {
+      "epoch": 0.6702069935931503,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7729,
+      "step": 13143
+    },
+    {
+      "epoch": 0.6702579870492557,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 13144
+    },
+    {
+      "epoch": 0.6703089805053611,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7512,
+      "step": 13145
+    },
+    {
+      "epoch": 0.6703599739614665,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 13146
+    },
+    {
+      "epoch": 0.6704109674175719,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7749,
+      "step": 13147
+    },
+    {
+      "epoch": 0.6704619608736773,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7784,
+      "step": 13148
+    },
+    {
+      "epoch": 0.6705129543297826,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7835,
+      "step": 13149
+    },
+    {
+      "epoch": 0.670563947785888,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6706149412419934,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 13151
+    },
+    {
+      "epoch": 0.6706659346980989,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 13152
+    },
+    {
+      "epoch": 0.6707169281542043,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 13153
+    },
+    {
+      "epoch": 0.6707679216103096,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 13154
+    },
+    {
+      "epoch": 0.670818915066415,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 13155
+    },
+    {
+      "epoch": 0.6708699085225204,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 13156
+    },
+    {
+      "epoch": 0.6709209019786257,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 13157
+    },
+    {
+      "epoch": 0.6709718954347311,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13158
+    },
+    {
+      "epoch": 0.6710228888908366,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7772,
+      "step": 13159
+    },
+    {
+      "epoch": 0.671073882346942,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6711248758030474,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 13161
+    },
+    {
+      "epoch": 0.6711758692591527,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 13162
+    },
+    {
+      "epoch": 0.6712268627152581,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 13163
+    },
+    {
+      "epoch": 0.6712778561713635,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 13164
+    },
+    {
+      "epoch": 0.6713288496274689,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 13165
+    },
+    {
+      "epoch": 0.6713798430835743,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 13166
+    },
+    {
+      "epoch": 0.6714308365396797,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 13167
+    },
+    {
+      "epoch": 0.6714818299957851,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 13168
+    },
+    {
+      "epoch": 0.6715328234518905,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13169
+    },
+    {
+      "epoch": 0.6715838169079958,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6716348103641012,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 13171
+    },
+    {
+      "epoch": 0.6716858038202066,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 13172
+    },
+    {
+      "epoch": 0.6717367972763121,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 13173
+    },
+    {
+      "epoch": 0.6717877907324175,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 13174
+    },
+    {
+      "epoch": 0.6718387841885228,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7801,
+      "step": 13175
+    },
+    {
+      "epoch": 0.6718897776446282,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 13176
+    },
+    {
+      "epoch": 0.6719407711007336,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 13177
+    },
+    {
+      "epoch": 0.671991764556839,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 13178
+    },
+    {
+      "epoch": 0.6720427580129443,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 13179
+    },
+    {
+      "epoch": 0.6720937514690498,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6721447449251552,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 13181
+    },
+    {
+      "epoch": 0.6721957383812606,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 13182
+    },
+    {
+      "epoch": 0.6722467318373659,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 13183
+    },
+    {
+      "epoch": 0.6722977252934713,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 13184
+    },
+    {
+      "epoch": 0.6723487187495767,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 13185
+    },
+    {
+      "epoch": 0.6723997122056821,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 13186
+    },
+    {
+      "epoch": 0.6724507056617876,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 13187
+    },
+    {
+      "epoch": 0.6725016991178929,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 13188
+    },
+    {
+      "epoch": 0.6725526925739983,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 13189
+    },
+    {
+      "epoch": 0.6726036860301037,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7847,
+      "step": 13190
+    },
+    {
+      "epoch": 0.672654679486209,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 13191
+    },
+    {
+      "epoch": 0.6727056729423144,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7895,
+      "step": 13192
+    },
+    {
+      "epoch": 0.6727566663984198,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7759,
+      "step": 13193
+    },
+    {
+      "epoch": 0.6728076598545253,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 13194
+    },
+    {
+      "epoch": 0.6728586533106307,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13195
+    },
+    {
+      "epoch": 0.672909646766736,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13196
+    },
+    {
+      "epoch": 0.6729606402228414,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 13197
+    },
+    {
+      "epoch": 0.6730116336789468,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 13198
+    },
+    {
+      "epoch": 0.6730626271350522,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 13199
+    },
+    {
+      "epoch": 0.6731136205911575,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7903,
+      "step": 13200
+    },
+    {
+      "epoch": 0.673164614047263,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 13201
+    },
+    {
+      "epoch": 0.6732156075033684,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 13202
+    },
+    {
+      "epoch": 0.6732666009594738,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 13203
+    },
+    {
+      "epoch": 0.6733175944155791,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.779,
+      "step": 13204
+    },
+    {
+      "epoch": 0.6733685878716845,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 13205
+    },
+    {
+      "epoch": 0.6734195813277899,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7697,
+      "step": 13206
+    },
+    {
+      "epoch": 0.6734705747838953,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 13207
+    },
+    {
+      "epoch": 0.6735215682400008,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 13208
+    },
+    {
+      "epoch": 0.6735725616961061,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 13209
+    },
+    {
+      "epoch": 0.6736235551522115,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6736745486083169,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 13211
+    },
+    {
+      "epoch": 0.6737255420644223,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 13212
+    },
+    {
+      "epoch": 0.6737765355205276,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 13213
+    },
+    {
+      "epoch": 0.673827528976633,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7775,
+      "step": 13214
+    },
+    {
+      "epoch": 0.6738785224327385,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13215
+    },
+    {
+      "epoch": 0.6739295158888439,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 13216
+    },
+    {
+      "epoch": 0.6739805093449492,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 13217
+    },
+    {
+      "epoch": 0.6740315028010546,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 13218
+    },
+    {
+      "epoch": 0.67408249625716,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 13219
+    },
+    {
+      "epoch": 0.6741334897132654,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6741844831693707,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 13221
+    },
+    {
+      "epoch": 0.6742354766254762,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 13222
+    },
+    {
+      "epoch": 0.6742864700815816,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 13223
+    },
+    {
+      "epoch": 0.674337463537687,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 13224
+    },
+    {
+      "epoch": 0.6743884569937924,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 13225
+    },
+    {
+      "epoch": 0.6744394504498977,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 13226
+    },
+    {
+      "epoch": 0.6744904439060031,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7893,
+      "step": 13227
+    },
+    {
+      "epoch": 0.6745414373621085,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 13228
+    },
+    {
+      "epoch": 0.674592430818214,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 13229
+    },
+    {
+      "epoch": 0.6746434242743193,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.788,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6746944177304247,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 13231
+    },
+    {
+      "epoch": 0.6747454111865301,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 13232
+    },
+    {
+      "epoch": 0.6747964046426355,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 13233
+    },
+    {
+      "epoch": 0.6748473980987408,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 13234
+    },
+    {
+      "epoch": 0.6748983915548462,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 13235
+    },
+    {
+      "epoch": 0.6749493850109517,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 13236
+    },
+    {
+      "epoch": 0.6750003784670571,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 13237
+    },
+    {
+      "epoch": 0.6750513719231624,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 13238
+    },
+    {
+      "epoch": 0.6751023653792678,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 13239
+    },
+    {
+      "epoch": 0.6751533588353732,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6752043522914786,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 13241
+    },
+    {
+      "epoch": 0.675255345747584,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 13242
+    },
+    {
+      "epoch": 0.6753063392036894,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 13243
+    },
+    {
+      "epoch": 0.6753573326597948,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13244
+    },
+    {
+      "epoch": 0.6754083261159002,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 13245
+    },
+    {
+      "epoch": 0.6754593195720056,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 13246
+    },
+    {
+      "epoch": 0.6755103130281109,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 13247
+    },
+    {
+      "epoch": 0.6755613064842163,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7859,
+      "step": 13248
+    },
+    {
+      "epoch": 0.6756122999403217,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 13249
+    },
+    {
+      "epoch": 0.6756632933964272,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 13250
+    },
+    {
+      "epoch": 0.6757142868525325,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 13251
+    },
+    {
+      "epoch": 0.6757652803086379,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13252
+    },
+    {
+      "epoch": 0.6758162737647433,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 13253
+    },
+    {
+      "epoch": 0.6758672672208487,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 13254
+    },
+    {
+      "epoch": 0.675918260676954,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 13255
+    },
+    {
+      "epoch": 0.6759692541330594,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13256
+    },
+    {
+      "epoch": 0.6760202475891649,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 13257
+    },
+    {
+      "epoch": 0.6760712410452703,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 13258
+    },
+    {
+      "epoch": 0.6761222345013757,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7753,
+      "step": 13259
+    },
+    {
+      "epoch": 0.676173227957481,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6762242214135864,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 13261
+    },
+    {
+      "epoch": 0.6762752148696918,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 13262
+    },
+    {
+      "epoch": 0.6763262083257972,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 13263
+    },
+    {
+      "epoch": 0.6763772017819026,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 13264
+    },
+    {
+      "epoch": 0.676428195238008,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7793,
+      "step": 13265
+    },
+    {
+      "epoch": 0.6764791886941134,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 13266
+    },
+    {
+      "epoch": 0.6765301821502188,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7808,
+      "step": 13267
+    },
+    {
+      "epoch": 0.6765811756063241,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 13268
+    },
+    {
+      "epoch": 0.6766321690624295,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 13269
+    },
+    {
+      "epoch": 0.6766831625185349,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.7823,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6767341559746404,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 13271
+    },
+    {
+      "epoch": 0.6767851494307457,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7749,
+      "step": 13272
+    },
+    {
+      "epoch": 0.6768361428868511,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 13273
+    },
+    {
+      "epoch": 0.6768871363429565,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7779,
+      "step": 13274
+    },
+    {
+      "epoch": 0.6769381297990619,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 13275
+    },
+    {
+      "epoch": 0.6769891232551672,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 13276
+    },
+    {
+      "epoch": 0.6770401167112726,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 13277
+    },
+    {
+      "epoch": 0.6770911101673781,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 13278
+    },
+    {
+      "epoch": 0.6771421036234835,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 13279
+    },
+    {
+      "epoch": 0.6771930970795889,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6772440905356942,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 13281
+    },
+    {
+      "epoch": 0.6772950839917996,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 13282
+    },
+    {
+      "epoch": 0.677346077447905,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7764,
+      "step": 13283
+    },
+    {
+      "epoch": 0.6773970709040104,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 13284
+    },
+    {
+      "epoch": 0.6774480643601158,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 13285
+    },
+    {
+      "epoch": 0.6774990578162212,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7729,
+      "step": 13286
+    },
+    {
+      "epoch": 0.6775500512723266,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.782,
+      "step": 13287
+    },
+    {
+      "epoch": 0.677601044728432,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 13288
+    },
+    {
+      "epoch": 0.6776520381845373,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 13289
+    },
+    {
+      "epoch": 0.6777030316406427,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6777540250967481,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 13291
+    },
+    {
+      "epoch": 0.6778050185528536,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 13292
+    },
+    {
+      "epoch": 0.677856012008959,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 13293
+    },
+    {
+      "epoch": 0.6779070054650643,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 13294
+    },
+    {
+      "epoch": 0.6779579989211697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7673,
+      "step": 13295
+    },
+    {
+      "epoch": 0.6780089923772751,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 13296
+    },
+    {
+      "epoch": 0.6780599858333805,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 13297
+    },
+    {
+      "epoch": 0.6781109792894858,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 13298
+    },
+    {
+      "epoch": 0.6781619727455912,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7626,
+      "step": 13299
+    },
+    {
+      "epoch": 0.6782129662016967,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6782639596578021,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7672,
+      "step": 13301
+    },
+    {
+      "epoch": 0.6783149531139074,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 13302
+    },
+    {
+      "epoch": 0.6783659465700128,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 13303
+    },
+    {
+      "epoch": 0.6784169400261182,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 13304
+    },
+    {
+      "epoch": 0.6784679334822236,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 13305
+    },
+    {
+      "epoch": 0.6785189269383289,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 13306
+    },
+    {
+      "epoch": 0.6785699203944344,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.781,
+      "step": 13307
+    },
+    {
+      "epoch": 0.6786209138505398,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7737,
+      "step": 13308
+    },
+    {
+      "epoch": 0.6786719073066452,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8277,
+      "step": 13309
+    },
+    {
+      "epoch": 0.6787229007627505,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6787738942188559,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 13311
+    },
+    {
+      "epoch": 0.6788248876749613,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7788,
+      "step": 13312
+    },
+    {
+      "epoch": 0.6788758811310667,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 13313
+    },
+    {
+      "epoch": 0.6789268745871722,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 13314
+    },
+    {
+      "epoch": 0.6789778680432775,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 13315
+    },
+    {
+      "epoch": 0.6790288614993829,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 13316
+    },
+    {
+      "epoch": 0.6790798549554883,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 13317
+    },
+    {
+      "epoch": 0.6791308484115937,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13318
+    },
+    {
+      "epoch": 0.679181841867699,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 13319
+    },
+    {
+      "epoch": 0.6792328353238044,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6792838287799099,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 13321
+    },
+    {
+      "epoch": 0.6793348222360153,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 13322
+    },
+    {
+      "epoch": 0.6793858156921206,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 13323
+    },
+    {
+      "epoch": 0.679436809148226,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 13324
+    },
+    {
+      "epoch": 0.6794878026043314,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 13325
+    },
+    {
+      "epoch": 0.6795387960604368,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 13326
+    },
+    {
+      "epoch": 0.6795897895165421,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 13327
+    },
+    {
+      "epoch": 0.6796407829726476,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 13328
+    },
+    {
+      "epoch": 0.679691776428753,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 13329
+    },
+    {
+      "epoch": 0.6797427698848584,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6797937633409638,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13331
+    },
+    {
+      "epoch": 0.6798447567970691,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 13332
+    },
+    {
+      "epoch": 0.6798957502531745,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 13333
+    },
+    {
+      "epoch": 0.6799467437092799,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8335,
+      "step": 13334
+    },
+    {
+      "epoch": 0.6799977371653854,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 13335
+    },
+    {
+      "epoch": 0.6800487306214907,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 13336
+    },
+    {
+      "epoch": 0.6800997240775961,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7797,
+      "step": 13337
+    },
+    {
+      "epoch": 0.6801507175337015,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7703,
+      "step": 13338
+    },
+    {
+      "epoch": 0.6802017109898069,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7834,
+      "step": 13339
+    },
+    {
+      "epoch": 0.6802527044459122,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6803036979020176,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 13341
+    },
+    {
+      "epoch": 0.6803546913581231,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 13342
+    },
+    {
+      "epoch": 0.6804056848142285,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7689,
+      "step": 13343
+    },
+    {
+      "epoch": 0.6804566782703338,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 13344
+    },
+    {
+      "epoch": 0.6805076717264392,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13345
+    },
+    {
+      "epoch": 0.6805586651825446,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13346
+    },
+    {
+      "epoch": 0.68060965863865,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 13347
+    },
+    {
+      "epoch": 0.6806606520947553,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 13348
+    },
+    {
+      "epoch": 0.6807116455508608,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 13349
+    },
+    {
+      "epoch": 0.6807626390069662,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6808136324630716,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13351
+    },
+    {
+      "epoch": 0.680864625919177,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13352
+    },
+    {
+      "epoch": 0.6809156193752823,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 13353
+    },
+    {
+      "epoch": 0.6809666128313877,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 13354
+    },
+    {
+      "epoch": 0.6810176062874931,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 13355
+    },
+    {
+      "epoch": 0.6810685997435986,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 13356
+    },
+    {
+      "epoch": 0.6811195931997039,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 13357
+    },
+    {
+      "epoch": 0.6811705866558093,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 13358
+    },
+    {
+      "epoch": 0.6812215801119147,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 13359
+    },
+    {
+      "epoch": 0.6812725735680201,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6813235670241254,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13361
+    },
+    {
+      "epoch": 0.6813745604802308,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 13362
+    },
+    {
+      "epoch": 0.6814255539363363,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 13363
+    },
+    {
+      "epoch": 0.6814765473924417,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 13364
+    },
+    {
+      "epoch": 0.681527540848547,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 13365
+    },
+    {
+      "epoch": 0.6815785343046524,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7779,
+      "step": 13366
+    },
+    {
+      "epoch": 0.6816295277607578,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 13367
+    },
+    {
+      "epoch": 0.6816805212168632,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 13368
+    },
+    {
+      "epoch": 0.6817315146729686,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 13369
+    },
+    {
+      "epoch": 0.681782508129074,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6818335015851794,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 13371
+    },
+    {
+      "epoch": 0.6818844950412848,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13372
+    },
+    {
+      "epoch": 0.6819354884973902,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 13373
+    },
+    {
+      "epoch": 0.6819864819534955,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7553,
+      "step": 13374
+    },
+    {
+      "epoch": 0.6820374754096009,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13375
+    },
+    {
+      "epoch": 0.6820884688657063,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 13376
+    },
+    {
+      "epoch": 0.6821394623218118,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 13377
+    },
+    {
+      "epoch": 0.6821904557779171,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 13378
+    },
+    {
+      "epoch": 0.6822414492340225,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 13379
+    },
+    {
+      "epoch": 0.6822924426901279,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6823434361462333,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 13381
+    },
+    {
+      "epoch": 0.6823944296023386,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 13382
+    },
+    {
+      "epoch": 0.682445423058444,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 13383
+    },
+    {
+      "epoch": 0.6824964165145495,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 13384
+    },
+    {
+      "epoch": 0.6825474099706549,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 13385
+    },
+    {
+      "epoch": 0.6825984034267603,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 13386
+    },
+    {
+      "epoch": 0.6826493968828656,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 13387
+    },
+    {
+      "epoch": 0.682700390338971,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 13388
+    },
+    {
+      "epoch": 0.6827513837950764,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7732,
+      "step": 13389
+    },
+    {
+      "epoch": 0.6828023772511818,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7775,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6828533707072872,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 13391
+    },
+    {
+      "epoch": 0.6829043641633926,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 13392
+    },
+    {
+      "epoch": 0.682955357619498,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13393
+    },
+    {
+      "epoch": 0.6830063510756034,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7852,
+      "step": 13394
+    },
+    {
+      "epoch": 0.6830573445317087,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 13395
+    },
+    {
+      "epoch": 0.6831083379878141,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 13396
+    },
+    {
+      "epoch": 0.6831593314439195,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 13397
+    },
+    {
+      "epoch": 0.683210324900025,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8346,
+      "step": 13398
+    },
+    {
+      "epoch": 0.6832613183561304,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.8269,
+      "step": 13399
+    },
+    {
+      "epoch": 0.6833123118122357,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6833633052683411,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7734,
+      "step": 13401
+    },
+    {
+      "epoch": 0.6834142987244465,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.784,
+      "step": 13402
+    },
+    {
+      "epoch": 0.6834652921805519,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 13403
+    },
+    {
+      "epoch": 0.6835162856366572,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 13404
+    },
+    {
+      "epoch": 0.6835672790927627,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7841,
+      "step": 13405
+    },
+    {
+      "epoch": 0.6836182725488681,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 13406
+    },
+    {
+      "epoch": 0.6836692660049735,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7817,
+      "step": 13407
+    },
+    {
+      "epoch": 0.6837202594610788,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 13408
+    },
+    {
+      "epoch": 0.6837712529171842,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 13409
+    },
+    {
+      "epoch": 0.6838222463732896,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 13410
+    },
+    {
+      "epoch": 0.683873239829395,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 13411
+    },
+    {
+      "epoch": 0.6839242332855004,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 13412
+    },
+    {
+      "epoch": 0.6839752267416058,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 13413
+    },
+    {
+      "epoch": 0.6840262201977112,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 13414
+    },
+    {
+      "epoch": 0.6840772136538166,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 13415
+    },
+    {
+      "epoch": 0.684128207109922,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 13416
+    },
+    {
+      "epoch": 0.6841792005660273,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 13417
+    },
+    {
+      "epoch": 0.6842301940221327,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 13418
+    },
+    {
+      "epoch": 0.6842811874782382,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 13419
+    },
+    {
+      "epoch": 0.6843321809343436,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6843831743904489,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 13421
+    },
+    {
+      "epoch": 0.6844341678465543,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 13422
+    },
+    {
+      "epoch": 0.6844851613026597,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 13423
+    },
+    {
+      "epoch": 0.6845361547587651,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7765,
+      "step": 13424
+    },
+    {
+      "epoch": 0.6845871482148704,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 13425
+    },
+    {
+      "epoch": 0.6846381416709759,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 13426
+    },
+    {
+      "epoch": 0.6846891351270813,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7725,
+      "step": 13427
+    },
+    {
+      "epoch": 0.6847401285831867,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 13428
+    },
+    {
+      "epoch": 0.684791122039292,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 13429
+    },
+    {
+      "epoch": 0.6848421154953974,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6848931089515028,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 13431
+    },
+    {
+      "epoch": 0.6849441024076082,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13432
+    },
+    {
+      "epoch": 0.6849950958637137,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8342,
+      "step": 13433
+    },
+    {
+      "epoch": 0.685046089319819,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13434
+    },
+    {
+      "epoch": 0.6850970827759244,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 13435
+    },
+    {
+      "epoch": 0.6851480762320298,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 13436
+    },
+    {
+      "epoch": 0.6851990696881352,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7798,
+      "step": 13437
+    },
+    {
+      "epoch": 0.6852500631442405,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 13438
+    },
+    {
+      "epoch": 0.6853010566003459,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 13439
+    },
+    {
+      "epoch": 0.6853520500564514,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6854030435125568,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 13441
+    },
+    {
+      "epoch": 0.6854540369686621,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 13442
+    },
+    {
+      "epoch": 0.6855050304247675,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 13443
+    },
+    {
+      "epoch": 0.6855560238808729,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 13444
+    },
+    {
+      "epoch": 0.6856070173369783,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 13445
+    },
+    {
+      "epoch": 0.6856580107930836,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 13446
+    },
+    {
+      "epoch": 0.6857090042491891,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 13447
+    },
+    {
+      "epoch": 0.6857599977052945,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 13448
+    },
+    {
+      "epoch": 0.6858109911613999,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 13449
+    },
+    {
+      "epoch": 0.6858619846175052,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6859129780736106,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 13451
+    },
+    {
+      "epoch": 0.685963971529716,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 13452
+    },
+    {
+      "epoch": 0.6860149649858214,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 13453
+    },
+    {
+      "epoch": 0.6860659584419269,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 13454
+    },
+    {
+      "epoch": 0.6861169518980322,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 13455
+    },
+    {
+      "epoch": 0.6861679453541376,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 13456
+    },
+    {
+      "epoch": 0.686218938810243,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7742,
+      "step": 13457
+    },
+    {
+      "epoch": 0.6862699322663484,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 13458
+    },
+    {
+      "epoch": 0.6863209257224537,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13459
+    },
+    {
+      "epoch": 0.6863719191785591,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6864229126346646,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 13461
+    },
+    {
+      "epoch": 0.68647390609077,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 13462
+    },
+    {
+      "epoch": 0.6865248995468753,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 13463
+    },
+    {
+      "epoch": 0.6865758930029807,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8408,
+      "step": 13464
+    },
+    {
+      "epoch": 0.6866268864590861,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 13465
+    },
+    {
+      "epoch": 0.6866778799151915,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 13466
+    },
+    {
+      "epoch": 0.6867288733712968,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 13467
+    },
+    {
+      "epoch": 0.6867798668274023,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 13468
+    },
+    {
+      "epoch": 0.6868308602835077,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13469
+    },
+    {
+      "epoch": 0.6868818537396131,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6869328471957185,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13471
+    },
+    {
+      "epoch": 0.6869838406518238,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 13472
+    },
+    {
+      "epoch": 0.6870348341079292,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 13473
+    },
+    {
+      "epoch": 0.6870858275640346,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 13474
+    },
+    {
+      "epoch": 0.6871368210201401,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 13475
+    },
+    {
+      "epoch": 0.6871878144762454,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 13476
+    },
+    {
+      "epoch": 0.6872388079323508,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 13477
+    },
+    {
+      "epoch": 0.6872898013884562,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 13478
+    },
+    {
+      "epoch": 0.6873407948445616,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 13479
+    },
+    {
+      "epoch": 0.6873917883006669,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6874427817567723,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 13481
+    },
+    {
+      "epoch": 0.6874937752128778,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7783,
+      "step": 13482
+    },
+    {
+      "epoch": 0.6875447686689832,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 13483
+    },
+    {
+      "epoch": 0.6875957621250885,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7572,
+      "step": 13484
+    },
+    {
+      "epoch": 0.6876467555811939,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 13485
+    },
+    {
+      "epoch": 0.6876977490372993,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 13486
+    },
+    {
+      "epoch": 0.6877487424934047,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 13487
+    },
+    {
+      "epoch": 0.68779973594951,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 13488
+    },
+    {
+      "epoch": 0.6878507294056155,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 13489
+    },
+    {
+      "epoch": 0.6879017228617209,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 13490
+    },
+    {
+      "epoch": 0.6879527163178263,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 13491
+    },
+    {
+      "epoch": 0.6880037097739317,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7719,
+      "step": 13492
+    },
+    {
+      "epoch": 0.688054703230037,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 13493
+    },
+    {
+      "epoch": 0.6881056966861424,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 13494
+    },
+    {
+      "epoch": 0.6881566901422478,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 13495
+    },
+    {
+      "epoch": 0.6882076835983533,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 13496
+    },
+    {
+      "epoch": 0.6882586770544586,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 13497
+    },
+    {
+      "epoch": 0.688309670510564,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 13498
+    },
+    {
+      "epoch": 0.6883606639666694,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 13499
+    },
+    {
+      "epoch": 0.6884116574227748,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6884626508788801,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 13501
+    },
+    {
+      "epoch": 0.6885136443349855,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 13502
+    },
+    {
+      "epoch": 0.688564637791091,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 13503
+    },
+    {
+      "epoch": 0.6886156312471964,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7681,
+      "step": 13504
+    },
+    {
+      "epoch": 0.6886666247033018,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 13505
+    },
+    {
+      "epoch": 0.6887176181594071,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 13506
+    },
+    {
+      "epoch": 0.6887686116155125,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 13507
+    },
+    {
+      "epoch": 0.6888196050716179,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13508
+    },
+    {
+      "epoch": 0.6888705985277233,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 13509
+    },
+    {
+      "epoch": 0.6889215919838287,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6889725854399341,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 13511
+    },
+    {
+      "epoch": 0.6890235788960395,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 13512
+    },
+    {
+      "epoch": 0.6890745723521449,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 13513
+    },
+    {
+      "epoch": 0.6891255658082502,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 13514
+    },
+    {
+      "epoch": 0.6891765592643556,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13515
+    },
+    {
+      "epoch": 0.689227552720461,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 13516
+    },
+    {
+      "epoch": 0.6892785461765665,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13517
+    },
+    {
+      "epoch": 0.6893295396326719,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 13518
+    },
+    {
+      "epoch": 0.6893805330887772,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 13519
+    },
+    {
+      "epoch": 0.6894315265448826,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 13520
+    },
+    {
+      "epoch": 0.689482520000988,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13521
+    },
+    {
+      "epoch": 0.6895335134570933,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7728,
+      "step": 13522
+    },
+    {
+      "epoch": 0.6895845069131987,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 13523
+    },
+    {
+      "epoch": 0.6896355003693042,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13524
+    },
+    {
+      "epoch": 0.6896864938254096,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 13525
+    },
+    {
+      "epoch": 0.689737487281515,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 13526
+    },
+    {
+      "epoch": 0.6897884807376203,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 13527
+    },
+    {
+      "epoch": 0.6898394741937257,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 13528
+    },
+    {
+      "epoch": 0.6898904676498311,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.772,
+      "step": 13529
+    },
+    {
+      "epoch": 0.6899414611059365,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 13530
+    },
+    {
+      "epoch": 0.689992454562042,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 13531
+    },
+    {
+      "epoch": 0.6900434480181473,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 13532
+    },
+    {
+      "epoch": 0.6900944414742527,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 13533
+    },
+    {
+      "epoch": 0.6901454349303581,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 13534
+    },
+    {
+      "epoch": 0.6901964283864634,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7731,
+      "step": 13535
+    },
+    {
+      "epoch": 0.6902474218425688,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 13536
+    },
+    {
+      "epoch": 0.6902984152986742,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 13537
+    },
+    {
+      "epoch": 0.6903494087547797,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13538
+    },
+    {
+      "epoch": 0.6904004022108851,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 13539
+    },
+    {
+      "epoch": 0.6904513956669904,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6905023891230958,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 13541
+    },
+    {
+      "epoch": 0.6905533825792012,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 13542
+    },
+    {
+      "epoch": 0.6906043760353066,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 13543
+    },
+    {
+      "epoch": 0.6906553694914119,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 13544
+    },
+    {
+      "epoch": 0.6907063629475174,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 13545
+    },
+    {
+      "epoch": 0.6907573564036228,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 13546
+    },
+    {
+      "epoch": 0.6908083498597282,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 13547
+    },
+    {
+      "epoch": 0.6908593433158335,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 13548
+    },
+    {
+      "epoch": 0.6909103367719389,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 13549
+    },
+    {
+      "epoch": 0.6909613302280443,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7776,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6910123236841497,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 13551
+    },
+    {
+      "epoch": 0.6910633171402552,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 13552
+    },
+    {
+      "epoch": 0.6911143105963605,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7747,
+      "step": 13553
+    },
+    {
+      "epoch": 0.6911653040524659,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7723,
+      "step": 13554
+    },
+    {
+      "epoch": 0.6912162975085713,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 13555
+    },
+    {
+      "epoch": 0.6912672909646767,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7763,
+      "step": 13556
+    },
+    {
+      "epoch": 0.691318284420782,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 13557
+    },
+    {
+      "epoch": 0.6913692778768874,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 13558
+    },
+    {
+      "epoch": 0.6914202713329929,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 13559
+    },
+    {
+      "epoch": 0.6914712647890983,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6915222582452036,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7686,
+      "step": 13561
+    },
+    {
+      "epoch": 0.691573251701309,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 13562
+    },
+    {
+      "epoch": 0.6916242451574144,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 13563
+    },
+    {
+      "epoch": 0.6916752386135198,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 13564
+    },
+    {
+      "epoch": 0.6917262320696251,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 13565
+    },
+    {
+      "epoch": 0.6917772255257306,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.775,
+      "step": 13566
+    },
+    {
+      "epoch": 0.691828218981836,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7793,
+      "step": 13567
+    },
+    {
+      "epoch": 0.6918792124379414,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 13568
+    },
+    {
+      "epoch": 0.6919302058940467,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 13569
+    },
+    {
+      "epoch": 0.6919811993501521,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6920321928062575,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 13571
+    },
+    {
+      "epoch": 0.6920831862623629,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 13572
+    },
+    {
+      "epoch": 0.6921341797184684,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 13573
+    },
+    {
+      "epoch": 0.6921851731745737,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 13574
+    },
+    {
+      "epoch": 0.6922361666306791,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 13575
+    },
+    {
+      "epoch": 0.6922871600867845,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7856,
+      "step": 13576
+    },
+    {
+      "epoch": 0.6923381535428899,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 13577
+    },
+    {
+      "epoch": 0.6923891469989952,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 13578
+    },
+    {
+      "epoch": 0.6924401404551006,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 13579
+    },
+    {
+      "epoch": 0.6924911339112061,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6925421273673115,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 13581
+    },
+    {
+      "epoch": 0.6925931208234168,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13582
+    },
+    {
+      "epoch": 0.6926441142795222,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 13583
+    },
+    {
+      "epoch": 0.6926951077356276,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 13584
+    },
+    {
+      "epoch": 0.692746101191733,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 13585
+    },
+    {
+      "epoch": 0.6927970946478383,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 13586
+    },
+    {
+      "epoch": 0.6928480881039438,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7716,
+      "step": 13587
+    },
+    {
+      "epoch": 0.6928990815600492,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 13588
+    },
+    {
+      "epoch": 0.6929500750161546,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13589
+    },
+    {
+      "epoch": 0.69300106847226,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7724,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6930520619283653,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 13591
+    },
+    {
+      "epoch": 0.6931030553844707,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8339,
+      "step": 13592
+    },
+    {
+      "epoch": 0.6931540488405761,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 13593
+    },
+    {
+      "epoch": 0.6932050422966816,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7646,
+      "step": 13594
+    },
+    {
+      "epoch": 0.6932560357527869,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 13595
+    },
+    {
+      "epoch": 0.6933070292088923,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 13596
+    },
+    {
+      "epoch": 0.6933580226649977,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 13597
+    },
+    {
+      "epoch": 0.6934090161211031,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7878,
+      "step": 13598
+    },
+    {
+      "epoch": 0.6934600095772084,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 13599
+    },
+    {
+      "epoch": 0.6935110030333138,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7738,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6935619964894193,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 13601
+    },
+    {
+      "epoch": 0.6936129899455247,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 13602
+    },
+    {
+      "epoch": 0.69366398340163,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 13603
+    },
+    {
+      "epoch": 0.6937149768577354,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 13604
+    },
+    {
+      "epoch": 0.6937659703138408,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 13605
+    },
+    {
+      "epoch": 0.6938169637699462,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 13606
+    },
+    {
+      "epoch": 0.6938679572260515,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 13607
+    },
+    {
+      "epoch": 0.693918950682157,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 13608
+    },
+    {
+      "epoch": 0.6939699441382624,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 13609
+    },
+    {
+      "epoch": 0.6940209375943678,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6940719310504732,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 13611
+    },
+    {
+      "epoch": 0.6941229245065785,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 13612
+    },
+    {
+      "epoch": 0.6941739179626839,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 13613
+    },
+    {
+      "epoch": 0.6942249114187893,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 13614
+    },
+    {
+      "epoch": 0.6942759048748948,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 13615
+    },
+    {
+      "epoch": 0.6943268983310001,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 13616
+    },
+    {
+      "epoch": 0.6943778917871055,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 13617
+    },
+    {
+      "epoch": 0.6944288852432109,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 13618
+    },
+    {
+      "epoch": 0.6944798786993163,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 13619
+    },
+    {
+      "epoch": 0.6945308721554216,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 13620
+    },
+    {
+      "epoch": 0.694581865611527,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 13621
+    },
+    {
+      "epoch": 0.6946328590676325,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 13622
+    },
+    {
+      "epoch": 0.6946838525237379,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13623
+    },
+    {
+      "epoch": 0.6947348459798433,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 13624
+    },
+    {
+      "epoch": 0.6947858394359486,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 13625
+    },
+    {
+      "epoch": 0.694836832892054,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 13626
+    },
+    {
+      "epoch": 0.6948878263481594,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 13627
+    },
+    {
+      "epoch": 0.6949388198042648,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 13628
+    },
+    {
+      "epoch": 0.6949898132603702,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 13629
+    },
+    {
+      "epoch": 0.6950408067164756,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 13630
+    },
+    {
+      "epoch": 0.695091800172581,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13631
+    },
+    {
+      "epoch": 0.6951427936286864,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7878,
+      "step": 13632
+    },
+    {
+      "epoch": 0.6951937870847917,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 13633
+    },
+    {
+      "epoch": 0.6952447805408971,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 13634
+    },
+    {
+      "epoch": 0.6952957739970025,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 13635
+    },
+    {
+      "epoch": 0.695346767453108,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 13636
+    },
+    {
+      "epoch": 0.6953977609092133,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 13637
+    },
+    {
+      "epoch": 0.6954487543653187,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 13638
+    },
+    {
+      "epoch": 0.6954997478214241,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 13639
+    },
+    {
+      "epoch": 0.6955507412775295,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7775,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6956017347336348,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 13641
+    },
+    {
+      "epoch": 0.6956527281897402,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 13642
+    },
+    {
+      "epoch": 0.6957037216458457,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 13643
+    },
+    {
+      "epoch": 0.6957547151019511,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7805,
+      "step": 13644
+    },
+    {
+      "epoch": 0.6958057085580565,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 13645
+    },
+    {
+      "epoch": 0.6958567020141618,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7776,
+      "step": 13646
+    },
+    {
+      "epoch": 0.6959076954702672,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 13647
+    },
+    {
+      "epoch": 0.6959586889263726,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 13648
+    },
+    {
+      "epoch": 0.696009682382478,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 13649
+    },
+    {
+      "epoch": 0.6960606758385834,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6961116692946888,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 13651
+    },
+    {
+      "epoch": 0.6961626627507942,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 13652
+    },
+    {
+      "epoch": 0.6962136562068996,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 13653
+    },
+    {
+      "epoch": 0.6962646496630049,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 13654
+    },
+    {
+      "epoch": 0.6963156431191103,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 13655
+    },
+    {
+      "epoch": 0.6963666365752157,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 13656
+    },
+    {
+      "epoch": 0.6964176300313212,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7806,
+      "step": 13657
+    },
+    {
+      "epoch": 0.6964686234874266,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 13658
+    },
+    {
+      "epoch": 0.6965196169435319,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 13659
+    },
+    {
+      "epoch": 0.6965706103996373,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6966216038557427,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 13661
+    },
+    {
+      "epoch": 0.696672597311848,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 13662
+    },
+    {
+      "epoch": 0.6967235907679534,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 13663
+    },
+    {
+      "epoch": 0.6967745842240589,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 13664
+    },
+    {
+      "epoch": 0.6968255776801643,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 13665
+    },
+    {
+      "epoch": 0.6968765711362697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 13666
+    },
+    {
+      "epoch": 0.696927564592375,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13667
+    },
+    {
+      "epoch": 0.6969785580484804,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 13668
+    },
+    {
+      "epoch": 0.6970295515045858,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 13669
+    },
+    {
+      "epoch": 0.6970805449606912,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6971315384167966,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7795,
+      "step": 13671
+    },
+    {
+      "epoch": 0.697182531872902,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 13672
+    },
+    {
+      "epoch": 0.6972335253290074,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 13673
+    },
+    {
+      "epoch": 0.6972845187851128,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 13674
+    },
+    {
+      "epoch": 0.6973355122412181,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 13675
+    },
+    {
+      "epoch": 0.6973865056973235,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 13676
+    },
+    {
+      "epoch": 0.6974374991534289,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 13677
+    },
+    {
+      "epoch": 0.6974884926095344,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 13678
+    },
+    {
+      "epoch": 0.6975394860656398,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 13679
+    },
+    {
+      "epoch": 0.6975904795217451,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6976414729778505,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 13681
+    },
+    {
+      "epoch": 0.6976924664339559,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 13682
+    },
+    {
+      "epoch": 0.6977434598900613,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 13683
+    },
+    {
+      "epoch": 0.6977944533461666,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 13684
+    },
+    {
+      "epoch": 0.6978454468022721,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 13685
+    },
+    {
+      "epoch": 0.6978964402583775,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 13686
+    },
+    {
+      "epoch": 0.6979474337144829,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 13687
+    },
+    {
+      "epoch": 0.6979984271705882,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13688
+    },
+    {
+      "epoch": 0.6980494206266936,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 13689
+    },
+    {
+      "epoch": 0.698100414082799,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6981514075389044,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 13691
+    },
+    {
+      "epoch": 0.6982024009950099,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7737,
+      "step": 13692
+    },
+    {
+      "epoch": 0.6982533944511152,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 13693
+    },
+    {
+      "epoch": 0.6983043879072206,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 13694
+    },
+    {
+      "epoch": 0.698355381363326,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 13695
+    },
+    {
+      "epoch": 0.6984063748194314,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8365,
+      "step": 13696
+    },
+    {
+      "epoch": 0.6984573682755367,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7984,
+      "step": 13697
+    },
+    {
+      "epoch": 0.6985083617316421,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 13698
+    },
+    {
+      "epoch": 0.6985593551877476,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7865,
+      "step": 13699
+    },
+    {
+      "epoch": 0.698610348643853,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6986613420999583,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 13701
+    },
+    {
+      "epoch": 0.6987123355560637,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 13702
+    },
+    {
+      "epoch": 0.6987633290121691,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.774,
+      "step": 13703
+    },
+    {
+      "epoch": 0.6988143224682745,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 13704
+    },
+    {
+      "epoch": 0.6988653159243798,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 13705
+    },
+    {
+      "epoch": 0.6989163093804853,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 13706
+    },
+    {
+      "epoch": 0.6989673028365907,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 13707
+    },
+    {
+      "epoch": 0.6990182962926961,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 13708
+    },
+    {
+      "epoch": 0.6990692897488014,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 13709
+    },
+    {
+      "epoch": 0.6991202832049068,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6991712766610122,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 13711
+    },
+    {
+      "epoch": 0.6992222701171176,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 13712
+    },
+    {
+      "epoch": 0.6992732635732231,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 13713
+    },
+    {
+      "epoch": 0.6993242570293284,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 13714
+    },
+    {
+      "epoch": 0.6993752504854338,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 13715
+    },
+    {
+      "epoch": 0.6994262439415392,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 13716
+    },
+    {
+      "epoch": 0.6994772373976446,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 13717
+    },
+    {
+      "epoch": 0.6995282308537499,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 13718
+    },
+    {
+      "epoch": 0.6995792243098553,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7814,
+      "step": 13719
+    },
+    {
+      "epoch": 0.6996302177659608,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6996812112220662,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 13721
+    },
+    {
+      "epoch": 0.6997322046781715,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 13722
+    },
+    {
+      "epoch": 0.6997831981342769,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 13723
+    },
+    {
+      "epoch": 0.6998341915903823,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 13724
+    },
+    {
+      "epoch": 0.6998851850464877,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 13725
+    },
+    {
+      "epoch": 0.699936178502593,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 13726
+    },
+    {
+      "epoch": 0.6999871719586985,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 13727
+    },
+    {
+      "epoch": 0.7000381654148039,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 13728
+    },
+    {
+      "epoch": 0.7000891588709093,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 13729
+    },
+    {
+      "epoch": 0.7001401523270147,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 13730
+    },
+    {
+      "epoch": 0.70019114578312,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 13731
+    },
+    {
+      "epoch": 0.7002421392392254,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13732
+    },
+    {
+      "epoch": 0.7002931326953308,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 13733
+    },
+    {
+      "epoch": 0.7003441261514363,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7854,
+      "step": 13734
+    },
+    {
+      "epoch": 0.7003951196075416,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 13735
+    },
+    {
+      "epoch": 0.700446113063647,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 13736
+    },
+    {
+      "epoch": 0.7004971065197524,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 13737
+    },
+    {
+      "epoch": 0.7005480999758578,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 13738
+    },
+    {
+      "epoch": 0.7005990934319631,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 13739
+    },
+    {
+      "epoch": 0.7006500868880685,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7633,
+      "step": 13740
+    },
+    {
+      "epoch": 0.700701080344174,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 13741
+    },
+    {
+      "epoch": 0.7007520738002794,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 13742
+    },
+    {
+      "epoch": 0.7008030672563847,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 13743
+    },
+    {
+      "epoch": 0.7008540607124901,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 13744
+    },
+    {
+      "epoch": 0.7009050541685955,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 13745
+    },
+    {
+      "epoch": 0.7009560476247009,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 13746
+    },
+    {
+      "epoch": 0.7010070410808062,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 13747
+    },
+    {
+      "epoch": 0.7010580345369117,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 13748
+    },
+    {
+      "epoch": 0.7011090279930171,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 13749
+    },
+    {
+      "epoch": 0.7011600214491225,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7012110149052279,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 13751
+    },
+    {
+      "epoch": 0.7012620083613332,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7815,
+      "step": 13752
+    },
+    {
+      "epoch": 0.7013130018174386,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 13753
+    },
+    {
+      "epoch": 0.701363995273544,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 13754
+    },
+    {
+      "epoch": 0.7014149887296495,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 13755
+    },
+    {
+      "epoch": 0.7014659821857548,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 13756
+    },
+    {
+      "epoch": 0.7015169756418602,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 13757
+    },
+    {
+      "epoch": 0.7015679690979656,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13758
+    },
+    {
+      "epoch": 0.701618962554071,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 13759
+    },
+    {
+      "epoch": 0.7016699560101763,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7017209494662817,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 13761
+    },
+    {
+      "epoch": 0.7017719429223872,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 13762
+    },
+    {
+      "epoch": 0.7018229363784926,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 13763
+    },
+    {
+      "epoch": 0.701873929834598,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 13764
+    },
+    {
+      "epoch": 0.7019249232907033,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 13765
+    },
+    {
+      "epoch": 0.7019759167468087,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 13766
+    },
+    {
+      "epoch": 0.7020269102029141,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 13767
+    },
+    {
+      "epoch": 0.7020779036590195,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 13768
+    },
+    {
+      "epoch": 0.7021288971151249,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 13769
+    },
+    {
+      "epoch": 0.7021798905712303,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7022308840273357,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13771
+    },
+    {
+      "epoch": 0.7022818774834411,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13772
+    },
+    {
+      "epoch": 0.7023328709395464,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 2.7749,
+      "step": 13773
+    },
+    {
+      "epoch": 0.7023838643956518,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13774
+    },
+    {
+      "epoch": 0.7024348578517572,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 13775
+    },
+    {
+      "epoch": 0.7024858513078627,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 13776
+    },
+    {
+      "epoch": 0.702536844763968,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 13777
+    },
+    {
+      "epoch": 0.7025878382200734,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 13778
+    },
+    {
+      "epoch": 0.7026388316761788,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 13779
+    },
+    {
+      "epoch": 0.7026898251322842,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7027408185883895,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 13781
+    },
+    {
+      "epoch": 0.7027918120444949,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 13782
+    },
+    {
+      "epoch": 0.7028428055006004,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 13783
+    },
+    {
+      "epoch": 0.7028937989567058,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7925,
+      "step": 13784
+    },
+    {
+      "epoch": 0.7029447924128112,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 13785
+    },
+    {
+      "epoch": 0.7029957858689165,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 13786
+    },
+    {
+      "epoch": 0.7030467793250219,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 13787
+    },
+    {
+      "epoch": 0.7030977727811273,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 13788
+    },
+    {
+      "epoch": 0.7031487662372327,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7588,
+      "step": 13789
+    },
+    {
+      "epoch": 0.7031997596933381,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7032507531494435,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 13791
+    },
+    {
+      "epoch": 0.7033017466055489,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13792
+    },
+    {
+      "epoch": 0.7033527400616543,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7692,
+      "step": 13793
+    },
+    {
+      "epoch": 0.7034037335177596,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 13794
+    },
+    {
+      "epoch": 0.703454726973865,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13795
+    },
+    {
+      "epoch": 0.7035057204299704,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 13796
+    },
+    {
+      "epoch": 0.7035567138860759,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 13797
+    },
+    {
+      "epoch": 0.7036077073421813,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7822,
+      "step": 13798
+    },
+    {
+      "epoch": 0.7036587007982866,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 13799
+    },
+    {
+      "epoch": 0.703709694254392,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7876,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7037606877104974,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 13801
+    },
+    {
+      "epoch": 0.7038116811666028,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 13802
+    },
+    {
+      "epoch": 0.7038626746227081,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 13803
+    },
+    {
+      "epoch": 0.7039136680788136,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7878,
+      "step": 13804
+    },
+    {
+      "epoch": 0.703964661534919,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 13805
+    },
+    {
+      "epoch": 0.7040156549910244,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7823,
+      "step": 13806
+    },
+    {
+      "epoch": 0.7040666484471297,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 13807
+    },
+    {
+      "epoch": 0.7041176419032351,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 13808
+    },
+    {
+      "epoch": 0.7041686353593405,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 13809
+    },
+    {
+      "epoch": 0.7042196288154459,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7042706222715513,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8211,
+      "step": 13811
+    },
+    {
+      "epoch": 0.7043216157276567,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 13812
+    },
+    {
+      "epoch": 0.7043726091837621,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 13813
+    },
+    {
+      "epoch": 0.7044236026398675,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 13814
+    },
+    {
+      "epoch": 0.7044745960959728,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7855,
+      "step": 13815
+    },
+    {
+      "epoch": 0.7045255895520782,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 13816
+    },
+    {
+      "epoch": 0.7045765830081836,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13817
+    },
+    {
+      "epoch": 0.7046275764642891,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 13818
+    },
+    {
+      "epoch": 0.7046785699203945,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 13819
+    },
+    {
+      "epoch": 0.7047295633764998,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7047805568326052,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 13821
+    },
+    {
+      "epoch": 0.7048315502887106,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 13822
+    },
+    {
+      "epoch": 0.704882543744816,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 13823
+    },
+    {
+      "epoch": 0.7049335372009213,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 13824
+    },
+    {
+      "epoch": 0.7049845306570268,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 13825
+    },
+    {
+      "epoch": 0.7050355241131322,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 13826
+    },
+    {
+      "epoch": 0.7050865175692376,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 13827
+    },
+    {
+      "epoch": 0.705137511025343,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 13828
+    },
+    {
+      "epoch": 0.7051885044814483,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 13829
+    },
+    {
+      "epoch": 0.7052394979375537,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7052904913936591,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13831
+    },
+    {
+      "epoch": 0.7053414848497646,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 13832
+    },
+    {
+      "epoch": 0.7053924783058699,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 13833
+    },
+    {
+      "epoch": 0.7054434717619753,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 13834
+    },
+    {
+      "epoch": 0.7054944652180807,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7811,
+      "step": 13835
+    },
+    {
+      "epoch": 0.705545458674186,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 13836
+    },
+    {
+      "epoch": 0.7055964521302914,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 13837
+    },
+    {
+      "epoch": 0.7056474455863968,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 13838
+    },
+    {
+      "epoch": 0.7056984390425023,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 13839
+    },
+    {
+      "epoch": 0.7057494324986077,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 13840
+    },
+    {
+      "epoch": 0.705800425954713,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 13841
+    },
+    {
+      "epoch": 0.7058514194108184,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 13842
+    },
+    {
+      "epoch": 0.7059024128669238,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 13843
+    },
+    {
+      "epoch": 0.7059534063230292,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 13844
+    },
+    {
+      "epoch": 0.7060043997791345,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 13845
+    },
+    {
+      "epoch": 0.70605539323524,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 13846
+    },
+    {
+      "epoch": 0.7061063866913454,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 13847
+    },
+    {
+      "epoch": 0.7061573801474508,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 13848
+    },
+    {
+      "epoch": 0.7062083736035561,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 13849
+    },
+    {
+      "epoch": 0.7062593670596615,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7063103605157669,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7878,
+      "step": 13851
+    },
+    {
+      "epoch": 0.7063613539718723,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 13852
+    },
+    {
+      "epoch": 0.7064123474279778,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 13853
+    },
+    {
+      "epoch": 0.7064633408840831,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8345,
+      "step": 13854
+    },
+    {
+      "epoch": 0.7065143343401885,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13855
+    },
+    {
+      "epoch": 0.7065653277962939,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 13856
+    },
+    {
+      "epoch": 0.7066163212523993,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 13857
+    },
+    {
+      "epoch": 0.7066673147085046,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 13858
+    },
+    {
+      "epoch": 0.70671830816461,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 13859
+    },
+    {
+      "epoch": 0.7067693016207155,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7068202950768209,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 13861
+    },
+    {
+      "epoch": 0.7068712885329262,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 13862
+    },
+    {
+      "epoch": 0.7069222819890316,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 13863
+    },
+    {
+      "epoch": 0.706973275445137,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 13864
+    },
+    {
+      "epoch": 0.7070242689012424,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8308,
+      "step": 13865
+    },
+    {
+      "epoch": 0.7070752623573477,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 13866
+    },
+    {
+      "epoch": 0.7071262558134532,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 13867
+    },
+    {
+      "epoch": 0.7071772492695586,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7703,
+      "step": 13868
+    },
+    {
+      "epoch": 0.707228242725664,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 13869
+    },
+    {
+      "epoch": 0.7072792361817694,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 13870
+    },
+    {
+      "epoch": 0.7073302296378747,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 13871
+    },
+    {
+      "epoch": 0.7073812230939801,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 13872
+    },
+    {
+      "epoch": 0.7074322165500855,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 13873
+    },
+    {
+      "epoch": 0.707483210006191,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 13874
+    },
+    {
+      "epoch": 0.7075342034622963,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7729,
+      "step": 13875
+    },
+    {
+      "epoch": 0.7075851969184017,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 13876
+    },
+    {
+      "epoch": 0.7076361903745071,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7658,
+      "step": 13877
+    },
+    {
+      "epoch": 0.7076871838306125,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 13878
+    },
+    {
+      "epoch": 0.7077381772867178,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 13879
+    },
+    {
+      "epoch": 0.7077891707428232,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 13880
+    },
+    {
+      "epoch": 0.7078401641989287,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 13881
+    },
+    {
+      "epoch": 0.7078911576550341,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7931,
+      "step": 13882
+    },
+    {
+      "epoch": 0.7079421511111395,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 13883
+    },
+    {
+      "epoch": 0.7079931445672448,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 13884
+    },
+    {
+      "epoch": 0.7080441380233502,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 13885
+    },
+    {
+      "epoch": 0.7080951314794556,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 13886
+    },
+    {
+      "epoch": 0.708146124935561,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 13887
+    },
+    {
+      "epoch": 0.7081971183916664,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 13888
+    },
+    {
+      "epoch": 0.7082481118477718,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 13889
+    },
+    {
+      "epoch": 0.7082991053038772,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 13890
+    },
+    {
+      "epoch": 0.7083500987599826,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 13891
+    },
+    {
+      "epoch": 0.7084010922160879,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7743,
+      "step": 13892
+    },
+    {
+      "epoch": 0.7084520856721933,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 13893
+    },
+    {
+      "epoch": 0.7085030791282987,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 13894
+    },
+    {
+      "epoch": 0.7085540725844042,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 13895
+    },
+    {
+      "epoch": 0.7086050660405095,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 13896
+    },
+    {
+      "epoch": 0.7086560594966149,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 13897
+    },
+    {
+      "epoch": 0.7087070529527203,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 13898
+    },
+    {
+      "epoch": 0.7087580464088257,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 13899
+    },
+    {
+      "epoch": 0.708809039864931,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7088600333210364,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 13901
+    },
+    {
+      "epoch": 0.7089110267771419,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 13902
+    },
+    {
+      "epoch": 0.7089620202332473,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 13903
+    },
+    {
+      "epoch": 0.7090130136893527,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 13904
+    },
+    {
+      "epoch": 0.709064007145458,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 13905
+    },
+    {
+      "epoch": 0.7091150006015634,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7706,
+      "step": 13906
+    },
+    {
+      "epoch": 0.7091659940576688,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 13907
+    },
+    {
+      "epoch": 0.7092169875137742,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 13908
+    },
+    {
+      "epoch": 0.7092679809698796,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13909
+    },
+    {
+      "epoch": 0.709318974425985,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13910
+    },
+    {
+      "epoch": 0.7093699678820904,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 13911
+    },
+    {
+      "epoch": 0.7094209613381958,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 13912
+    },
+    {
+      "epoch": 0.7094719547943011,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 13913
+    },
+    {
+      "epoch": 0.7095229482504065,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 13914
+    },
+    {
+      "epoch": 0.7095739417065119,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7917,
+      "step": 13915
+    },
+    {
+      "epoch": 0.7096249351626174,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8456,
+      "step": 13916
+    },
+    {
+      "epoch": 0.7096759286187228,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 13917
+    },
+    {
+      "epoch": 0.7097269220748281,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 13918
+    },
+    {
+      "epoch": 0.7097779155309335,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 13919
+    },
+    {
+      "epoch": 0.7098289089870389,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 13920
+    },
+    {
+      "epoch": 0.7098799024431443,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 13921
+    },
+    {
+      "epoch": 0.7099308958992496,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 13922
+    },
+    {
+      "epoch": 0.7099818893553551,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 13923
+    },
+    {
+      "epoch": 0.7100328828114605,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 13924
+    },
+    {
+      "epoch": 0.7100838762675659,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 13925
+    },
+    {
+      "epoch": 0.7101348697236712,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 13926
+    },
+    {
+      "epoch": 0.7101858631797766,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 13927
+    },
+    {
+      "epoch": 0.710236856635882,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8105,
+      "step": 13928
+    },
+    {
+      "epoch": 0.7102878500919874,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 13929
+    },
+    {
+      "epoch": 0.7103388435480928,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 13930
+    },
+    {
+      "epoch": 0.7103898370041982,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 13931
+    },
+    {
+      "epoch": 0.7104408304603036,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 13932
+    },
+    {
+      "epoch": 0.710491823916409,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 13933
+    },
+    {
+      "epoch": 0.7105428173725143,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 13934
+    },
+    {
+      "epoch": 0.7105938108286197,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 13935
+    },
+    {
+      "epoch": 0.7106448042847251,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 13936
+    },
+    {
+      "epoch": 0.7106957977408306,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 13937
+    },
+    {
+      "epoch": 0.710746791196936,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 13938
+    },
+    {
+      "epoch": 0.7107977846530413,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 13939
+    },
+    {
+      "epoch": 0.7108487781091467,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 13940
+    },
+    {
+      "epoch": 0.7108997715652521,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 13941
+    },
+    {
+      "epoch": 0.7109507650213575,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8258,
+      "step": 13942
+    },
+    {
+      "epoch": 0.7110017584774628,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 13943
+    },
+    {
+      "epoch": 0.7110527519335683,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 13944
+    },
+    {
+      "epoch": 0.7111037453896737,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 13945
+    },
+    {
+      "epoch": 0.7111547388457791,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7696,
+      "step": 13946
+    },
+    {
+      "epoch": 0.7112057323018844,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7938,
+      "step": 13947
+    },
+    {
+      "epoch": 0.7112567257579898,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 13948
+    },
+    {
+      "epoch": 0.7113077192140952,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 13949
+    },
+    {
+      "epoch": 0.7113587126702006,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 13950
+    },
+    {
+      "epoch": 0.711409706126306,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8497,
+      "step": 13951
+    },
+    {
+      "epoch": 0.7114606995824114,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.765,
+      "step": 13952
+    },
+    {
+      "epoch": 0.7115116930385168,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 13953
+    },
+    {
+      "epoch": 0.7115626864946222,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 13954
+    },
+    {
+      "epoch": 0.7116136799507276,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 13955
+    },
+    {
+      "epoch": 0.7116646734068329,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 13956
+    },
+    {
+      "epoch": 0.7117156668629383,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 13957
+    },
+    {
+      "epoch": 0.7117666603190438,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 13958
+    },
+    {
+      "epoch": 0.7118176537751492,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 13959
+    },
+    {
+      "epoch": 0.7118686472312545,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7682,
+      "step": 13960
+    },
+    {
+      "epoch": 0.7119196406873599,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7858,
+      "step": 13961
+    },
+    {
+      "epoch": 0.7119706341434653,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 13962
+    },
+    {
+      "epoch": 0.7120216275995707,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 13963
+    },
+    {
+      "epoch": 0.712072621055676,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7879,
+      "step": 13964
+    },
+    {
+      "epoch": 0.7121236145117815,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 13965
+    },
+    {
+      "epoch": 0.7121746079678869,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 13966
+    },
+    {
+      "epoch": 0.7122256014239923,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 13967
+    },
+    {
+      "epoch": 0.7122765948800976,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 13968
+    },
+    {
+      "epoch": 0.712327588336203,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.788,
+      "step": 13969
+    },
+    {
+      "epoch": 0.7123785817923084,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 13970
+    },
+    {
+      "epoch": 0.7124295752484138,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 13971
+    },
+    {
+      "epoch": 0.7124805687045193,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 13972
+    },
+    {
+      "epoch": 0.7125315621606246,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7903,
+      "step": 13973
+    },
+    {
+      "epoch": 0.71258255561673,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 13974
+    },
+    {
+      "epoch": 0.7126335490728354,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 13975
+    },
+    {
+      "epoch": 0.7126845425289408,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7752,
+      "step": 13976
+    },
+    {
+      "epoch": 0.7127355359850461,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7883,
+      "step": 13977
+    },
+    {
+      "epoch": 0.7127865294411515,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 13978
+    },
+    {
+      "epoch": 0.712837522897257,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 13979
+    },
+    {
+      "epoch": 0.7128885163533624,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 13980
+    },
+    {
+      "epoch": 0.7129395098094677,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 13981
+    },
+    {
+      "epoch": 0.7129905032655731,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 13982
+    },
+    {
+      "epoch": 0.7130414967216785,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 13983
+    },
+    {
+      "epoch": 0.7130924901777839,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 13984
+    },
+    {
+      "epoch": 0.7131434836338892,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 13985
+    },
+    {
+      "epoch": 0.7131944770899947,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 13986
+    },
+    {
+      "epoch": 0.7132454705461001,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7874,
+      "step": 13987
+    },
+    {
+      "epoch": 0.7132964640022055,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 13988
+    },
+    {
+      "epoch": 0.7133474574583109,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 13989
+    },
+    {
+      "epoch": 0.7133984509144162,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 13990
+    },
+    {
+      "epoch": 0.7134494443705216,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 13991
+    },
+    {
+      "epoch": 0.713500437826627,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 13992
+    },
+    {
+      "epoch": 0.7135514312827325,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 13993
+    },
+    {
+      "epoch": 0.7136024247388378,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 13994
+    },
+    {
+      "epoch": 0.7136534181949432,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 13995
+    },
+    {
+      "epoch": 0.7137044116510486,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 13996
+    },
+    {
+      "epoch": 0.713755405107154,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 13997
+    },
+    {
+      "epoch": 0.7138063985632593,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 13998
+    },
+    {
+      "epoch": 0.7138573920193647,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 13999
+    },
+    {
+      "epoch": 0.7139083854754702,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 14000
+    },
+    {
+      "epoch": 0.7139593789315756,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 14001
+    },
+    {
+      "epoch": 0.714010372387681,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 14002
+    },
+    {
+      "epoch": 0.7140613658437863,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7773,
+      "step": 14003
+    },
+    {
+      "epoch": 0.7141123592998917,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14004
+    },
+    {
+      "epoch": 0.7141633527559971,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 14005
+    },
+    {
+      "epoch": 0.7142143462121024,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7749,
+      "step": 14006
+    },
+    {
+      "epoch": 0.7142653396682079,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 14007
+    },
+    {
+      "epoch": 0.7143163331243133,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 14008
+    },
+    {
+      "epoch": 0.7143673265804187,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 14009
+    },
+    {
+      "epoch": 0.7144183200365241,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 14010
+    },
+    {
+      "epoch": 0.7144693134926294,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 14011
+    },
+    {
+      "epoch": 0.7145203069487348,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 14012
+    },
+    {
+      "epoch": 0.7145713004048402,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 14013
+    },
+    {
+      "epoch": 0.7146222938609457,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 14014
+    },
+    {
+      "epoch": 0.714673287317051,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 14015
+    },
+    {
+      "epoch": 0.7147242807731564,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14016
+    },
+    {
+      "epoch": 0.7147752742292618,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.7694,
+      "step": 14017
+    },
+    {
+      "epoch": 0.7148262676853672,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 14018
+    },
+    {
+      "epoch": 0.7148772611414725,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 14019
+    },
+    {
+      "epoch": 0.7149282545975779,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 14020
+    },
+    {
+      "epoch": 0.7149792480536834,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 14021
+    },
+    {
+      "epoch": 0.7150302415097888,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 14022
+    },
+    {
+      "epoch": 0.7150812349658942,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 14023
+    },
+    {
+      "epoch": 0.7151322284219995,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 14024
+    },
+    {
+      "epoch": 0.7151832218781049,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 14025
+    },
+    {
+      "epoch": 0.7152342153342103,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8465,
+      "step": 14026
+    },
+    {
+      "epoch": 0.7152852087903157,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 14027
+    },
+    {
+      "epoch": 0.7153362022464211,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 14028
+    },
+    {
+      "epoch": 0.7153871957025265,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 14029
+    },
+    {
+      "epoch": 0.7154381891586319,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 14030
+    },
+    {
+      "epoch": 0.7154891826147373,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 14031
+    },
+    {
+      "epoch": 0.7155401760708426,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7793,
+      "step": 14032
+    },
+    {
+      "epoch": 0.715591169526948,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 14033
+    },
+    {
+      "epoch": 0.7156421629830534,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 14034
+    },
+    {
+      "epoch": 0.7156931564391589,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7665,
+      "step": 14035
+    },
+    {
+      "epoch": 0.7157441498952642,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 14036
+    },
+    {
+      "epoch": 0.7157951433513696,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 14037
+    },
+    {
+      "epoch": 0.715846136807475,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 14038
+    },
+    {
+      "epoch": 0.7158971302635804,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 14039
+    },
+    {
+      "epoch": 0.7159481237196857,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7782,
+      "step": 14040
+    },
+    {
+      "epoch": 0.7159991171757911,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7861,
+      "step": 14041
+    },
+    {
+      "epoch": 0.7160501106318966,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 14042
+    },
+    {
+      "epoch": 0.716101104088002,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 14043
+    },
+    {
+      "epoch": 0.7161520975441074,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 14044
+    },
+    {
+      "epoch": 0.7162030910002127,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 14045
+    },
+    {
+      "epoch": 0.7162540844563181,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 14046
+    },
+    {
+      "epoch": 0.7163050779124235,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 14047
+    },
+    {
+      "epoch": 0.7163560713685289,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 14048
+    },
+    {
+      "epoch": 0.7164070648246343,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 14049
+    },
+    {
+      "epoch": 0.7164580582807397,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 14050
+    },
+    {
+      "epoch": 0.7165090517368451,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 14051
+    },
+    {
+      "epoch": 0.7165600451929505,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 14052
+    },
+    {
+      "epoch": 0.7166110386490558,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 14053
+    },
+    {
+      "epoch": 0.7166620321051612,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 14054
+    },
+    {
+      "epoch": 0.7167130255612666,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 14055
+    },
+    {
+      "epoch": 0.7167640190173721,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 14056
+    },
+    {
+      "epoch": 0.7168150124734775,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 14057
+    },
+    {
+      "epoch": 0.7168660059295828,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 14058
+    },
+    {
+      "epoch": 0.7169169993856882,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7755,
+      "step": 14059
+    },
+    {
+      "epoch": 0.7169679928417936,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 14060
+    },
+    {
+      "epoch": 0.717018986297899,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7816,
+      "step": 14061
+    },
+    {
+      "epoch": 0.7170699797540043,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8392,
+      "step": 14062
+    },
+    {
+      "epoch": 0.7171209732101098,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8266,
+      "step": 14063
+    },
+    {
+      "epoch": 0.7171719666662152,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 14064
+    },
+    {
+      "epoch": 0.7172229601223206,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8356,
+      "step": 14065
+    },
+    {
+      "epoch": 0.7172739535784259,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 14066
+    },
+    {
+      "epoch": 0.7173249470345313,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 14067
+    },
+    {
+      "epoch": 0.7173759404906367,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 14068
+    },
+    {
+      "epoch": 0.7174269339467421,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 14069
+    },
+    {
+      "epoch": 0.7174779274028475,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8362,
+      "step": 14070
+    },
+    {
+      "epoch": 0.7175289208589529,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 14071
+    },
+    {
+      "epoch": 0.7175799143150583,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 14072
+    },
+    {
+      "epoch": 0.7176309077711637,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 14073
+    },
+    {
+      "epoch": 0.717681901227269,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 14074
+    },
+    {
+      "epoch": 0.7177328946833744,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 14075
+    },
+    {
+      "epoch": 0.7177838881394798,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 14076
+    },
+    {
+      "epoch": 0.7178348815955853,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 14077
+    },
+    {
+      "epoch": 0.7178858750516907,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 14078
+    },
+    {
+      "epoch": 0.717936868507796,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 14079
+    },
+    {
+      "epoch": 0.7179878619639014,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7934,
+      "step": 14080
+    },
+    {
+      "epoch": 0.7180388554200068,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 14081
+    },
+    {
+      "epoch": 0.7180898488761122,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8429,
+      "step": 14082
+    },
+    {
+      "epoch": 0.7181408423322175,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 14083
+    },
+    {
+      "epoch": 0.718191835788323,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 14084
+    },
+    {
+      "epoch": 0.7182428292444284,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 14085
+    },
+    {
+      "epoch": 0.7182938227005338,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 14086
+    },
+    {
+      "epoch": 0.7183448161566391,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 14087
+    },
+    {
+      "epoch": 0.7183958096127445,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 14088
+    },
+    {
+      "epoch": 0.7184468030688499,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 14089
+    },
+    {
+      "epoch": 0.7184977965249553,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7848,
+      "step": 14090
+    },
+    {
+      "epoch": 0.7185487899810608,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 14091
+    },
+    {
+      "epoch": 0.7185997834371661,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 14092
+    },
+    {
+      "epoch": 0.7186507768932715,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14093
+    },
+    {
+      "epoch": 0.7187017703493769,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14094
+    },
+    {
+      "epoch": 0.7187527638054823,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 14095
+    },
+    {
+      "epoch": 0.7188037572615876,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 14096
+    },
+    {
+      "epoch": 0.718854750717693,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14097
+    },
+    {
+      "epoch": 0.7189057441737985,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 14098
+    },
+    {
+      "epoch": 0.7189567376299039,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7701,
+      "step": 14099
+    },
+    {
+      "epoch": 0.7190077310860092,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7190587245421146,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 14101
+    },
+    {
+      "epoch": 0.71910971799822,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 14102
+    },
+    {
+      "epoch": 0.7191607114543254,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 14103
+    },
+    {
+      "epoch": 0.7192117049104307,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7957,
+      "step": 14104
+    },
+    {
+      "epoch": 0.7192626983665362,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 14105
+    },
+    {
+      "epoch": 0.7193136918226416,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14106
+    },
+    {
+      "epoch": 0.719364685278747,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 14107
+    },
+    {
+      "epoch": 0.7194156787348523,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 14108
+    },
+    {
+      "epoch": 0.7194666721909577,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 14109
+    },
+    {
+      "epoch": 0.7195176656470631,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 14110
+    },
+    {
+      "epoch": 0.7195686591031685,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14111
+    },
+    {
+      "epoch": 0.719619652559274,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8242,
+      "step": 14112
+    },
+    {
+      "epoch": 0.7196706460153793,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8248,
+      "step": 14113
+    },
+    {
+      "epoch": 0.7197216394714847,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 14114
+    },
+    {
+      "epoch": 0.7197726329275901,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8415,
+      "step": 14115
+    },
+    {
+      "epoch": 0.7198236263836955,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7654,
+      "step": 14116
+    },
+    {
+      "epoch": 0.7198746198398008,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 14117
+    },
+    {
+      "epoch": 0.7199256132959062,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 14118
+    },
+    {
+      "epoch": 0.7199766067520117,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 14119
+    },
+    {
+      "epoch": 0.7200276002081171,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7200785936642224,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 14121
+    },
+    {
+      "epoch": 0.7201295871203278,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 14122
+    },
+    {
+      "epoch": 0.7201805805764332,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 14123
+    },
+    {
+      "epoch": 0.7202315740325386,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 14124
+    },
+    {
+      "epoch": 0.7202825674886439,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 14125
+    },
+    {
+      "epoch": 0.7203335609447494,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 14126
+    },
+    {
+      "epoch": 0.7203845544008548,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 14127
+    },
+    {
+      "epoch": 0.7204355478569602,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14128
+    },
+    {
+      "epoch": 0.7204865413130656,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 14129
+    },
+    {
+      "epoch": 0.7205375347691709,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 14130
+    },
+    {
+      "epoch": 0.7205885282252763,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 14131
+    },
+    {
+      "epoch": 0.7206395216813817,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8338,
+      "step": 14132
+    },
+    {
+      "epoch": 0.7206905151374872,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 14133
+    },
+    {
+      "epoch": 0.7207415085935925,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 14134
+    },
+    {
+      "epoch": 0.7207925020496979,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 14135
+    },
+    {
+      "epoch": 0.7208434955058033,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 14136
+    },
+    {
+      "epoch": 0.7208944889619087,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 14137
+    },
+    {
+      "epoch": 0.720945482418014,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 14138
+    },
+    {
+      "epoch": 0.7209964758741194,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 14139
+    },
+    {
+      "epoch": 0.7210474693302249,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7210984627863303,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 14141
+    },
+    {
+      "epoch": 0.7211494562424356,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14142
+    },
+    {
+      "epoch": 0.721200449698541,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 14143
+    },
+    {
+      "epoch": 0.7212514431546464,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 14144
+    },
+    {
+      "epoch": 0.7213024366107518,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 14145
+    },
+    {
+      "epoch": 0.7213534300668571,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7804,
+      "step": 14146
+    },
+    {
+      "epoch": 0.7214044235229626,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8394,
+      "step": 14147
+    },
+    {
+      "epoch": 0.721455416979068,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14148
+    },
+    {
+      "epoch": 0.7215064104351734,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 14149
+    },
+    {
+      "epoch": 0.7215574038912788,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 14150
+    },
+    {
+      "epoch": 0.7216083973473841,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 14151
+    },
+    {
+      "epoch": 0.7216593908034895,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 14152
+    },
+    {
+      "epoch": 0.7217103842595949,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.7787,
+      "step": 14153
+    },
+    {
+      "epoch": 0.7217613777157004,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 14154
+    },
+    {
+      "epoch": 0.7218123711718057,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14155
+    },
+    {
+      "epoch": 0.7218633646279111,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14156
+    },
+    {
+      "epoch": 0.7219143580840165,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 14157
+    },
+    {
+      "epoch": 0.7219653515401219,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 14158
+    },
+    {
+      "epoch": 0.7220163449962272,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 14159
+    },
+    {
+      "epoch": 0.7220673384523326,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7221183319084381,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 14161
+    },
+    {
+      "epoch": 0.7221693253645435,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 14162
+    },
+    {
+      "epoch": 0.7222203188206489,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 14163
+    },
+    {
+      "epoch": 0.7222713122767542,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 14164
+    },
+    {
+      "epoch": 0.7223223057328596,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 14165
+    },
+    {
+      "epoch": 0.722373299188965,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 14166
+    },
+    {
+      "epoch": 0.7224242926450704,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 14167
+    },
+    {
+      "epoch": 0.7224752861011758,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 14168
+    },
+    {
+      "epoch": 0.7225262795572812,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 14169
+    },
+    {
+      "epoch": 0.7225772730133866,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 14170
+    },
+    {
+      "epoch": 0.722628266469492,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 14171
+    },
+    {
+      "epoch": 0.7226792599255973,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7888,
+      "step": 14172
+    },
+    {
+      "epoch": 0.7227302533817027,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 14173
+    },
+    {
+      "epoch": 0.7227812468378081,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7912,
+      "step": 14174
+    },
+    {
+      "epoch": 0.7228322402939136,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8354,
+      "step": 14175
+    },
+    {
+      "epoch": 0.722883233750019,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 14176
+    },
+    {
+      "epoch": 0.7229342272061243,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 14177
+    },
+    {
+      "epoch": 0.7229852206622297,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 14178
+    },
+    {
+      "epoch": 0.7230362141183351,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 14179
+    },
+    {
+      "epoch": 0.7230872075744404,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7231382010305458,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 14181
+    },
+    {
+      "epoch": 0.7231891944866513,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8502,
+      "step": 14182
+    },
+    {
+      "epoch": 0.7232401879427567,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 14183
+    },
+    {
+      "epoch": 0.7232911813988621,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 14184
+    },
+    {
+      "epoch": 0.7233421748549674,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 14185
+    },
+    {
+      "epoch": 0.7233931683110728,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 14186
+    },
+    {
+      "epoch": 0.7234441617671782,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 14187
+    },
+    {
+      "epoch": 0.7234951552232836,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 14188
+    },
+    {
+      "epoch": 0.723546148679389,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7853,
+      "step": 14189
+    },
+    {
+      "epoch": 0.7235971421354944,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7236481355915998,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 14191
+    },
+    {
+      "epoch": 0.7236991290477052,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 14192
+    },
+    {
+      "epoch": 0.7237501225038105,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 14193
+    },
+    {
+      "epoch": 0.7238011159599159,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 14194
+    },
+    {
+      "epoch": 0.7238521094160213,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7843,
+      "step": 14195
+    },
+    {
+      "epoch": 0.7239031028721268,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 14196
+    },
+    {
+      "epoch": 0.7239540963282322,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 14197
+    },
+    {
+      "epoch": 0.7240050897843375,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 14198
+    },
+    {
+      "epoch": 0.7240560832404429,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 14199
+    },
+    {
+      "epoch": 0.7241070766965483,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7241580701526537,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 14201
+    },
+    {
+      "epoch": 0.724209063608759,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14202
+    },
+    {
+      "epoch": 0.7242600570648645,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 14203
+    },
+    {
+      "epoch": 0.7243110505209699,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14204
+    },
+    {
+      "epoch": 0.7243620439770753,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 14205
+    },
+    {
+      "epoch": 0.7244130374331806,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 14206
+    },
+    {
+      "epoch": 0.724464030889286,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 14207
+    },
+    {
+      "epoch": 0.7245150243453914,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 14208
+    },
+    {
+      "epoch": 0.7245660178014968,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 14209
+    },
+    {
+      "epoch": 0.7246170112576023,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7246680047137076,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 14211
+    },
+    {
+      "epoch": 0.724718998169813,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 14212
+    },
+    {
+      "epoch": 0.7247699916259184,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 14213
+    },
+    {
+      "epoch": 0.7248209850820237,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 14214
+    },
+    {
+      "epoch": 0.7248719785381291,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 14215
+    },
+    {
+      "epoch": 0.7249229719942345,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 14216
+    },
+    {
+      "epoch": 0.72497396545034,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 14217
+    },
+    {
+      "epoch": 0.7250249589064454,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 14218
+    },
+    {
+      "epoch": 0.7250759523625507,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 14219
+    },
+    {
+      "epoch": 0.7251269458186561,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7251779392747615,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 14221
+    },
+    {
+      "epoch": 0.7252289327308669,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7849,
+      "step": 14222
+    },
+    {
+      "epoch": 0.7252799261869722,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 14223
+    },
+    {
+      "epoch": 0.7253309196430777,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 14224
+    },
+    {
+      "epoch": 0.7253819130991831,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 14225
+    },
+    {
+      "epoch": 0.7254329065552885,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14226
+    },
+    {
+      "epoch": 0.7254839000113938,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 14227
+    },
+    {
+      "epoch": 0.7255348934674992,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 14228
+    },
+    {
+      "epoch": 0.7255858869236046,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 14229
+    },
+    {
+      "epoch": 0.72563688037971,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 14230
+    },
+    {
+      "epoch": 0.7256878738358155,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 14231
+    },
+    {
+      "epoch": 0.7257388672919208,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 14232
+    },
+    {
+      "epoch": 0.7257898607480262,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 14233
+    },
+    {
+      "epoch": 0.7258408542041316,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 14234
+    },
+    {
+      "epoch": 0.725891847660237,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 14235
+    },
+    {
+      "epoch": 0.7259428411163423,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7793,
+      "step": 14236
+    },
+    {
+      "epoch": 0.7259938345724477,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 14237
+    },
+    {
+      "epoch": 0.7260448280285532,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 14238
+    },
+    {
+      "epoch": 0.7260958214846586,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 14239
+    },
+    {
+      "epoch": 0.7261468149407639,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7261978083968693,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.838,
+      "step": 14241
+    },
+    {
+      "epoch": 0.7262488018529747,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 14242
+    },
+    {
+      "epoch": 0.7262997953090801,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.821,
+      "step": 14243
+    },
+    {
+      "epoch": 0.7263507887651854,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 14244
+    },
+    {
+      "epoch": 0.7264017822212909,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 14245
+    },
+    {
+      "epoch": 0.7264527756773963,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.794,
+      "step": 14246
+    },
+    {
+      "epoch": 0.7265037691335017,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 14247
+    },
+    {
+      "epoch": 0.726554762589607,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.002,
+      "loss": 2.8333,
+      "step": 14248
+    },
+    {
+      "epoch": 0.7266057560457124,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 14249
+    },
+    {
+      "epoch": 0.7266567495018178,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7267077429579232,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7826,
+      "step": 14251
+    },
+    {
+      "epoch": 0.7267587364140287,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14252
+    },
+    {
+      "epoch": 0.726809729870134,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 14253
+    },
+    {
+      "epoch": 0.7268607233262394,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 14254
+    },
+    {
+      "epoch": 0.7269117167823448,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8037,
+      "step": 14255
+    },
+    {
+      "epoch": 0.7269627102384502,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 14256
+    },
+    {
+      "epoch": 0.7270137036945555,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 14257
+    },
+    {
+      "epoch": 0.7270646971506609,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 14258
+    },
+    {
+      "epoch": 0.7271156906067664,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 14259
+    },
+    {
+      "epoch": 0.7271666840628718,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7272176775189771,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 14261
+    },
+    {
+      "epoch": 0.7272686709750825,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8229,
+      "step": 14262
+    },
+    {
+      "epoch": 0.7273196644311879,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14263
+    },
+    {
+      "epoch": 0.7273706578872933,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7832,
+      "step": 14264
+    },
+    {
+      "epoch": 0.7274216513433986,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7831,
+      "step": 14265
+    },
+    {
+      "epoch": 0.7274726447995041,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 14266
+    },
+    {
+      "epoch": 0.7275236382556095,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7899,
+      "step": 14267
+    },
+    {
+      "epoch": 0.7275746317117149,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 14268
+    },
+    {
+      "epoch": 0.7276256251678203,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 14269
+    },
+    {
+      "epoch": 0.7276766186239256,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 14270
+    },
+    {
+      "epoch": 0.727727612080031,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14271
+    },
+    {
+      "epoch": 0.7277786055361364,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 14272
+    },
+    {
+      "epoch": 0.7278295989922419,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 14273
+    },
+    {
+      "epoch": 0.7278805924483472,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7854,
+      "step": 14274
+    },
+    {
+      "epoch": 0.7279315859044526,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8026,
+      "step": 14275
+    },
+    {
+      "epoch": 0.727982579360558,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 14276
+    },
+    {
+      "epoch": 0.7280335728166634,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 14277
+    },
+    {
+      "epoch": 0.7280845662727687,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 14278
+    },
+    {
+      "epoch": 0.7281355597288741,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 14279
+    },
+    {
+      "epoch": 0.7281865531849796,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 14280
+    },
+    {
+      "epoch": 0.728237546641085,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 14281
+    },
+    {
+      "epoch": 0.7282885400971904,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 14282
+    },
+    {
+      "epoch": 0.7283395335532957,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14283
+    },
+    {
+      "epoch": 0.7283905270094011,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8216,
+      "step": 14284
+    },
+    {
+      "epoch": 0.7284415204655065,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 14285
+    },
+    {
+      "epoch": 0.7284925139216119,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 14286
+    },
+    {
+      "epoch": 0.7285435073777173,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 14287
+    },
+    {
+      "epoch": 0.7285945008338227,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 14288
+    },
+    {
+      "epoch": 0.7286454942899281,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 14289
+    },
+    {
+      "epoch": 0.7286964877460335,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7716,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7287474812021388,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 14291
+    },
+    {
+      "epoch": 0.7287984746582442,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 14292
+    },
+    {
+      "epoch": 0.7288494681143496,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8406,
+      "step": 14293
+    },
+    {
+      "epoch": 0.7289004615704551,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 14294
+    },
+    {
+      "epoch": 0.7289514550265604,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 14295
+    },
+    {
+      "epoch": 0.7290024484826658,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 14296
+    },
+    {
+      "epoch": 0.7290534419387712,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 14297
+    },
+    {
+      "epoch": 0.7291044353948766,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8398,
+      "step": 14298
+    },
+    {
+      "epoch": 0.729155428850982,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 14299
+    },
+    {
+      "epoch": 0.7292064223070873,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7851,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7292574157631928,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 14301
+    },
+    {
+      "epoch": 0.7293084092192982,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14302
+    },
+    {
+      "epoch": 0.7293594026754036,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 14303
+    },
+    {
+      "epoch": 0.7294103961315089,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7893,
+      "step": 14304
+    },
+    {
+      "epoch": 0.7294613895876143,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 14305
+    },
+    {
+      "epoch": 0.7295123830437197,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 14306
+    },
+    {
+      "epoch": 0.7295633764998251,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8364,
+      "step": 14307
+    },
+    {
+      "epoch": 0.7296143699559305,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 14308
+    },
+    {
+      "epoch": 0.7296653634120359,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 14309
+    },
+    {
+      "epoch": 0.7297163568681413,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 14310
+    },
+    {
+      "epoch": 0.7297673503242467,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 14311
+    },
+    {
+      "epoch": 0.729818343780352,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7996,
+      "step": 14312
+    },
+    {
+      "epoch": 0.7298693372364574,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 14313
+    },
+    {
+      "epoch": 0.7299203306925628,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 14314
+    },
+    {
+      "epoch": 0.7299713241486683,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 14315
+    },
+    {
+      "epoch": 0.7300223176047737,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14316
+    },
+    {
+      "epoch": 0.730073311060879,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 14317
+    },
+    {
+      "epoch": 0.7301243045169844,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 14318
+    },
+    {
+      "epoch": 0.7301752979730898,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 14319
+    },
+    {
+      "epoch": 0.7302262914291952,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 14320
+    },
+    {
+      "epoch": 0.7302772848853005,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 14321
+    },
+    {
+      "epoch": 0.730328278341406,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 14322
+    },
+    {
+      "epoch": 0.7303792717975114,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 14323
+    },
+    {
+      "epoch": 0.7304302652536168,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 14324
+    },
+    {
+      "epoch": 0.7304812587097221,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8294,
+      "step": 14325
+    },
+    {
+      "epoch": 0.7305322521658275,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 14326
+    },
+    {
+      "epoch": 0.7305832456219329,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14327
+    },
+    {
+      "epoch": 0.7306342390780383,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 14328
+    },
+    {
+      "epoch": 0.7306852325341437,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 14329
+    },
+    {
+      "epoch": 0.7307362259902491,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7307872194463545,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 14331
+    },
+    {
+      "epoch": 0.7308382129024599,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 14332
+    },
+    {
+      "epoch": 0.7308892063585652,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14333
+    },
+    {
+      "epoch": 0.7309401998146706,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14334
+    },
+    {
+      "epoch": 0.730991193270776,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 14335
+    },
+    {
+      "epoch": 0.7310421867268815,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 14336
+    },
+    {
+      "epoch": 0.7310931801829869,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.783,
+      "step": 14337
+    },
+    {
+      "epoch": 0.7311441736390922,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 14338
+    },
+    {
+      "epoch": 0.7311951670951976,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 14339
+    },
+    {
+      "epoch": 0.731246160551303,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 14340
+    },
+    {
+      "epoch": 0.7312971540074084,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 14341
+    },
+    {
+      "epoch": 0.7313481474635137,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 14342
+    },
+    {
+      "epoch": 0.7313991409196192,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 14343
+    },
+    {
+      "epoch": 0.7314501343757246,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 14344
+    },
+    {
+      "epoch": 0.73150112783183,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14345
+    },
+    {
+      "epoch": 0.7315521212879353,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 14346
+    },
+    {
+      "epoch": 0.7316031147440407,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 14347
+    },
+    {
+      "epoch": 0.7316541082001461,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 14348
+    },
+    {
+      "epoch": 0.7317051016562515,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 14349
+    },
+    {
+      "epoch": 0.731756095112357,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7318070885684623,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 14351
+    },
+    {
+      "epoch": 0.7318580820245677,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 14352
+    },
+    {
+      "epoch": 0.7319090754806731,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 14353
+    },
+    {
+      "epoch": 0.7319600689367785,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8084,
+      "step": 14354
+    },
+    {
+      "epoch": 0.7320110623928838,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 14355
+    },
+    {
+      "epoch": 0.7320620558489892,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 14356
+    },
+    {
+      "epoch": 0.7321130493050947,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 14357
+    },
+    {
+      "epoch": 0.7321640427612001,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 14358
+    },
+    {
+      "epoch": 0.7322150362173054,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 14359
+    },
+    {
+      "epoch": 0.7322660296734108,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7323170231295162,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 14361
+    },
+    {
+      "epoch": 0.7323680165856216,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 14362
+    },
+    {
+      "epoch": 0.7324190100417269,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 14363
+    },
+    {
+      "epoch": 0.7324700034978324,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 14364
+    },
+    {
+      "epoch": 0.7325209969539378,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 14365
+    },
+    {
+      "epoch": 0.7325719904100432,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 14366
+    },
+    {
+      "epoch": 0.7326229838661485,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8551,
+      "step": 14367
+    },
+    {
+      "epoch": 0.7326739773222539,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8157,
+      "step": 14368
+    },
+    {
+      "epoch": 0.7327249707783593,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 14369
+    },
+    {
+      "epoch": 0.7327759642344647,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7328269576905702,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 14371
+    },
+    {
+      "epoch": 0.7328779511466755,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14372
+    },
+    {
+      "epoch": 0.7329289446027809,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 14373
+    },
+    {
+      "epoch": 0.7329799380588863,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 14374
+    },
+    {
+      "epoch": 0.7330309315149917,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 14375
+    },
+    {
+      "epoch": 0.733081924971097,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 14376
+    },
+    {
+      "epoch": 0.7331329184272024,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 14377
+    },
+    {
+      "epoch": 0.7331839118833079,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 14378
+    },
+    {
+      "epoch": 0.7332349053394133,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 14379
+    },
+    {
+      "epoch": 0.7332858987955186,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 14380
+    },
+    {
+      "epoch": 0.733336892251624,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 14381
+    },
+    {
+      "epoch": 0.7333878857077294,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 14382
+    },
+    {
+      "epoch": 0.7334388791638348,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8057,
+      "step": 14383
+    },
+    {
+      "epoch": 0.7334898726199401,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 14384
+    },
+    {
+      "epoch": 0.7335408660760456,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 14385
+    },
+    {
+      "epoch": 0.733591859532151,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 14386
+    },
+    {
+      "epoch": 0.7336428529882564,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 14387
+    },
+    {
+      "epoch": 0.7336938464443618,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 14388
+    },
+    {
+      "epoch": 0.7337448399004671,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7738,
+      "step": 14389
+    },
+    {
+      "epoch": 0.7337958333565725,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 14390
+    },
+    {
+      "epoch": 0.7338468268126779,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 14391
+    },
+    {
+      "epoch": 0.7338978202687834,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 14392
+    },
+    {
+      "epoch": 0.7339488137248887,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 14393
+    },
+    {
+      "epoch": 0.7339998071809941,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 14394
+    },
+    {
+      "epoch": 0.7340508006370995,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7839,
+      "step": 14395
+    },
+    {
+      "epoch": 0.7341017940932049,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 14396
+    },
+    {
+      "epoch": 0.7341527875493102,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8272,
+      "step": 14397
+    },
+    {
+      "epoch": 0.7342037810054156,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7651,
+      "step": 14398
+    },
+    {
+      "epoch": 0.7342547744615211,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14399
+    },
+    {
+      "epoch": 0.7343057679176265,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7343567613737318,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 14401
+    },
+    {
+      "epoch": 0.7344077548298372,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 14402
+    },
+    {
+      "epoch": 0.7344587482859426,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 14403
+    },
+    {
+      "epoch": 0.734509741742048,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 14404
+    },
+    {
+      "epoch": 0.7345607351981533,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.789,
+      "step": 14405
+    },
+    {
+      "epoch": 0.7346117286542588,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 14406
+    },
+    {
+      "epoch": 0.7346627221103642,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7776,
+      "step": 14407
+    },
+    {
+      "epoch": 0.7347137155664696,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7702,
+      "step": 14408
+    },
+    {
+      "epoch": 0.734764709022575,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8432,
+      "step": 14409
+    },
+    {
+      "epoch": 0.7348157024786803,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 14410
+    },
+    {
+      "epoch": 0.7348666959347857,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 14411
+    },
+    {
+      "epoch": 0.7349176893908911,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14412
+    },
+    {
+      "epoch": 0.7349686828469966,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7828,
+      "step": 14413
+    },
+    {
+      "epoch": 0.7350196763031019,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 14414
+    },
+    {
+      "epoch": 0.7350706697592073,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 14415
+    },
+    {
+      "epoch": 0.7351216632153127,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14416
+    },
+    {
+      "epoch": 0.7351726566714181,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 14417
+    },
+    {
+      "epoch": 0.7352236501275234,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8046,
+      "step": 14418
+    },
+    {
+      "epoch": 0.7352746435836288,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 14419
+    },
+    {
+      "epoch": 0.7353256370397343,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8483,
+      "step": 14420
+    },
+    {
+      "epoch": 0.7353766304958397,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14421
+    },
+    {
+      "epoch": 0.735427623951945,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 14422
+    },
+    {
+      "epoch": 0.7354786174080504,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 14423
+    },
+    {
+      "epoch": 0.7355296108641558,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 14424
+    },
+    {
+      "epoch": 0.7355806043202612,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 14425
+    },
+    {
+      "epoch": 0.7356315977763666,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 14426
+    },
+    {
+      "epoch": 0.735682591232472,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14427
+    },
+    {
+      "epoch": 0.7357335846885774,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 14428
+    },
+    {
+      "epoch": 0.7357845781446828,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 14429
+    },
+    {
+      "epoch": 0.7358355716007882,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 14430
+    },
+    {
+      "epoch": 0.7358865650568935,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.781,
+      "step": 14431
+    },
+    {
+      "epoch": 0.7359375585129989,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 14432
+    },
+    {
+      "epoch": 0.7359885519691043,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 14433
+    },
+    {
+      "epoch": 0.7360395454252098,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8461,
+      "step": 14434
+    },
+    {
+      "epoch": 0.7360905388813151,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 14435
+    },
+    {
+      "epoch": 0.7361415323374205,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 14436
+    },
+    {
+      "epoch": 0.7361925257935259,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 14437
+    },
+    {
+      "epoch": 0.7362435192496313,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 14438
+    },
+    {
+      "epoch": 0.7362945127057366,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 14439
+    },
+    {
+      "epoch": 0.736345506161842,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7754,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7363964996179475,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8251,
+      "step": 14441
+    },
+    {
+      "epoch": 0.7364474930740529,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8137,
+      "step": 14442
+    },
+    {
+      "epoch": 0.7364984865301583,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 14443
+    },
+    {
+      "epoch": 0.7365494799862636,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8134,
+      "step": 14444
+    },
+    {
+      "epoch": 0.736600473442369,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 14445
+    },
+    {
+      "epoch": 0.7366514668984744,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8053,
+      "step": 14446
+    },
+    {
+      "epoch": 0.7367024603545798,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8293,
+      "step": 14447
+    },
+    {
+      "epoch": 0.7367534538106852,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 14448
+    },
+    {
+      "epoch": 0.7368044472667906,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7687,
+      "step": 14449
+    },
+    {
+      "epoch": 0.736855440722896,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7369064341790014,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 14451
+    },
+    {
+      "epoch": 0.7369574276351067,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 14452
+    },
+    {
+      "epoch": 0.7370084210912121,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 14453
+    },
+    {
+      "epoch": 0.7370594145473175,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 14454
+    },
+    {
+      "epoch": 0.737110408003423,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 14455
+    },
+    {
+      "epoch": 0.7371614014595284,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.79,
+      "step": 14456
+    },
+    {
+      "epoch": 0.7372123949156337,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 14457
+    },
+    {
+      "epoch": 0.7372633883717391,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.787,
+      "step": 14458
+    },
+    {
+      "epoch": 0.7373143818278445,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 14459
+    },
+    {
+      "epoch": 0.7373653752839499,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7374163687400552,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 14461
+    },
+    {
+      "epoch": 0.7374673621961607,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8002,
+      "step": 14462
+    },
+    {
+      "epoch": 0.7375183556522661,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.7846,
+      "step": 14463
+    },
+    {
+      "epoch": 0.7375693491083715,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 14464
+    },
+    {
+      "epoch": 0.7376203425644768,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 14465
+    },
+    {
+      "epoch": 0.7376713360205822,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 14466
+    },
+    {
+      "epoch": 0.7377223294766876,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8286,
+      "step": 14467
+    },
+    {
+      "epoch": 0.737773322932793,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 14468
+    },
+    {
+      "epoch": 0.7378243163888984,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14469
+    },
+    {
+      "epoch": 0.7378753098450038,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7379263033011092,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 14471
+    },
+    {
+      "epoch": 0.7379772967572146,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14472
+    },
+    {
+      "epoch": 0.73802829021332,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 14473
+    },
+    {
+      "epoch": 0.7380792836694253,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 14474
+    },
+    {
+      "epoch": 0.7381302771255307,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 14475
+    },
+    {
+      "epoch": 0.7381812705816362,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 14476
+    },
+    {
+      "epoch": 0.7382322640377416,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 14477
+    },
+    {
+      "epoch": 0.7382832574938469,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8359,
+      "step": 14478
+    },
+    {
+      "epoch": 0.7383342509499523,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 14479
+    },
+    {
+      "epoch": 0.7383852444060577,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7384362378621631,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 14481
+    },
+    {
+      "epoch": 0.7384872313182684,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 14482
+    },
+    {
+      "epoch": 0.7385382247743739,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14483
+    },
+    {
+      "epoch": 0.7385892182304793,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8446,
+      "step": 14484
+    },
+    {
+      "epoch": 0.7386402116865847,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 14485
+    },
+    {
+      "epoch": 0.73869120514269,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7989,
+      "step": 14486
+    },
+    {
+      "epoch": 0.7387421985987954,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7871,
+      "step": 14487
+    },
+    {
+      "epoch": 0.7387931920549008,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 14488
+    },
+    {
+      "epoch": 0.7388441855110062,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 14489
+    },
+    {
+      "epoch": 0.7388951789671117,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 14490
+    },
+    {
+      "epoch": 0.738946172423217,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 14491
+    },
+    {
+      "epoch": 0.7389971658793224,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 14492
+    },
+    {
+      "epoch": 0.7390481593354278,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7876,
+      "step": 14493
+    },
+    {
+      "epoch": 0.7390991527915332,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14494
+    },
+    {
+      "epoch": 0.7391501462476385,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 14495
+    },
+    {
+      "epoch": 0.7392011397037439,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.78,
+      "step": 14496
+    },
+    {
+      "epoch": 0.7392521331598494,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 14497
+    },
+    {
+      "epoch": 0.7393031266159548,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 14498
+    },
+    {
+      "epoch": 0.7393541200720601,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14499
+    },
+    {
+      "epoch": 0.7394051135281655,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7394561069842709,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 14501
+    },
+    {
+      "epoch": 0.7395071004403763,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14502
+    },
+    {
+      "epoch": 0.7395580938964816,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7825,
+      "step": 14503
+    },
+    {
+      "epoch": 0.7396090873525871,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 14504
+    },
+    {
+      "epoch": 0.7396600808086925,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 14505
+    },
+    {
+      "epoch": 0.7397110742647979,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 14506
+    },
+    {
+      "epoch": 0.7397620677209032,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 14507
+    },
+    {
+      "epoch": 0.7398130611770086,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 14508
+    },
+    {
+      "epoch": 0.739864054633114,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 14509
+    },
+    {
+      "epoch": 0.7399150480892194,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14510
+    },
+    {
+      "epoch": 0.7399660415453249,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14511
+    },
+    {
+      "epoch": 0.7400170350014302,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 14512
+    },
+    {
+      "epoch": 0.7400680284575356,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7771,
+      "step": 14513
+    },
+    {
+      "epoch": 0.740119021913641,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 14514
+    },
+    {
+      "epoch": 0.7401700153697464,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.7921,
+      "step": 14515
+    },
+    {
+      "epoch": 0.7402210088258517,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 14516
+    },
+    {
+      "epoch": 0.7402720022819571,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 14517
+    },
+    {
+      "epoch": 0.7403229957380626,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 14518
+    },
+    {
+      "epoch": 0.740373989194168,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 14519
+    },
+    {
+      "epoch": 0.7404249826502733,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 14520
+    },
+    {
+      "epoch": 0.7404759761063787,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 14521
+    },
+    {
+      "epoch": 0.7405269695624841,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 14522
+    },
+    {
+      "epoch": 0.7405779630185895,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 14523
+    },
+    {
+      "epoch": 0.7406289564746948,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 14524
+    },
+    {
+      "epoch": 0.7406799499308003,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8407,
+      "step": 14525
+    },
+    {
+      "epoch": 0.7407309433869057,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7942,
+      "step": 14526
+    },
+    {
+      "epoch": 0.7407819368430111,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7656,
+      "step": 14527
+    },
+    {
+      "epoch": 0.7408329302991165,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 14528
+    },
+    {
+      "epoch": 0.7408839237552218,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 14529
+    },
+    {
+      "epoch": 0.7409349172113272,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 14530
+    },
+    {
+      "epoch": 0.7409859106674326,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 14531
+    },
+    {
+      "epoch": 0.7410369041235381,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 14532
+    },
+    {
+      "epoch": 0.7410878975796434,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 14533
+    },
+    {
+      "epoch": 0.7411388910357488,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 14534
+    },
+    {
+      "epoch": 0.7411898844918542,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.833,
+      "step": 14535
+    },
+    {
+      "epoch": 0.7412408779479596,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8166,
+      "step": 14536
+    },
+    {
+      "epoch": 0.7412918714040649,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 14537
+    },
+    {
+      "epoch": 0.7413428648601703,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 14538
+    },
+    {
+      "epoch": 0.7413938583162758,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 14539
+    },
+    {
+      "epoch": 0.7414448517723812,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 14540
+    },
+    {
+      "epoch": 0.7414958452284865,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7845,
+      "step": 14541
+    },
+    {
+      "epoch": 0.7415468386845919,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 14542
+    },
+    {
+      "epoch": 0.7415978321406973,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 14543
+    },
+    {
+      "epoch": 0.7416488255968027,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14544
+    },
+    {
+      "epoch": 0.741699819052908,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 14545
+    },
+    {
+      "epoch": 0.7417508125090135,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 14546
+    },
+    {
+      "epoch": 0.7418018059651189,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 14547
+    },
+    {
+      "epoch": 0.7418527994212243,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 14548
+    },
+    {
+      "epoch": 0.7419037928773297,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 14549
+    },
+    {
+      "epoch": 0.741954786333435,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8079,
+      "step": 14550
+    },
+    {
+      "epoch": 0.7420057797895404,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 14551
+    },
+    {
+      "epoch": 0.7420567732456458,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14552
+    },
+    {
+      "epoch": 0.7421077667017513,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 14553
+    },
+    {
+      "epoch": 0.7421587601578566,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 14554
+    },
+    {
+      "epoch": 0.742209753613962,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 14555
+    },
+    {
+      "epoch": 0.7422607470700674,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 14556
+    },
+    {
+      "epoch": 0.7423117405261728,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14557
+    },
+    {
+      "epoch": 0.7423627339822781,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8131,
+      "step": 14558
+    },
+    {
+      "epoch": 0.7424137274383835,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7881,
+      "step": 14559
+    },
+    {
+      "epoch": 0.742464720894489,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7864,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7425157143505944,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8025,
+      "step": 14561
+    },
+    {
+      "epoch": 0.7425667078066998,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8234,
+      "step": 14562
+    },
+    {
+      "epoch": 0.7426177012628051,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 14563
+    },
+    {
+      "epoch": 0.7426686947189105,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 14564
+    },
+    {
+      "epoch": 0.7427196881750159,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 14565
+    },
+    {
+      "epoch": 0.7427706816311213,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 14566
+    },
+    {
+      "epoch": 0.7428216750872267,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 14567
+    },
+    {
+      "epoch": 0.7428726685433321,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7757,
+      "step": 14568
+    },
+    {
+      "epoch": 0.7429236619994375,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 14569
+    },
+    {
+      "epoch": 0.7429746554555429,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7430256489116482,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 14571
+    },
+    {
+      "epoch": 0.7430766423677536,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 14572
+    },
+    {
+      "epoch": 0.743127635823859,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 14573
+    },
+    {
+      "epoch": 0.7431786292799645,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 14574
+    },
+    {
+      "epoch": 0.7432296227360699,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 14575
+    },
+    {
+      "epoch": 0.7432806161921752,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 14576
+    },
+    {
+      "epoch": 0.7433316096482806,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 14577
+    },
+    {
+      "epoch": 0.743382603104386,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14578
+    },
+    {
+      "epoch": 0.7434335965604913,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 14579
+    },
+    {
+      "epoch": 0.7434845900165967,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7435355834727022,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 14581
+    },
+    {
+      "epoch": 0.7435865769288076,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 14582
+    },
+    {
+      "epoch": 0.743637570384913,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7987,
+      "step": 14583
+    },
+    {
+      "epoch": 0.7436885638410183,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 14584
+    },
+    {
+      "epoch": 0.7437395572971237,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 14585
+    },
+    {
+      "epoch": 0.7437905507532291,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 14586
+    },
+    {
+      "epoch": 0.7438415442093345,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 14587
+    },
+    {
+      "epoch": 0.74389253766544,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 14588
+    },
+    {
+      "epoch": 0.7439435311215453,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8473,
+      "step": 14589
+    },
+    {
+      "epoch": 0.7439945245776507,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.796,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7440455180337561,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 14591
+    },
+    {
+      "epoch": 0.7440965114898614,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 14592
+    },
+    {
+      "epoch": 0.7441475049459668,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 14593
+    },
+    {
+      "epoch": 0.7441984984020722,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 14594
+    },
+    {
+      "epoch": 0.7442494918581777,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8052,
+      "step": 14595
+    },
+    {
+      "epoch": 0.7443004853142831,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 14596
+    },
+    {
+      "epoch": 0.7443514787703884,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 14597
+    },
+    {
+      "epoch": 0.7444024722264938,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 14598
+    },
+    {
+      "epoch": 0.7444534656825992,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14599
+    },
+    {
+      "epoch": 0.7445044591387046,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7445554525948099,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.7988,
+      "step": 14601
+    },
+    {
+      "epoch": 0.7446064460509154,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 14602
+    },
+    {
+      "epoch": 0.7446574395070208,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 14603
+    },
+    {
+      "epoch": 0.7447084329631262,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 14604
+    },
+    {
+      "epoch": 0.7447594264192315,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 14605
+    },
+    {
+      "epoch": 0.7448104198753369,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 14606
+    },
+    {
+      "epoch": 0.7448614133314423,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7922,
+      "step": 14607
+    },
+    {
+      "epoch": 0.7449124067875477,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 14608
+    },
+    {
+      "epoch": 0.7449634002436532,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 14609
+    },
+    {
+      "epoch": 0.7450143936997585,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7450653871558639,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 14611
+    },
+    {
+      "epoch": 0.7451163806119693,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 14612
+    },
+    {
+      "epoch": 0.7451673740680747,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 14613
+    },
+    {
+      "epoch": 0.74521836752418,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 14614
+    },
+    {
+      "epoch": 0.7452693609802854,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 14615
+    },
+    {
+      "epoch": 0.7453203544363909,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 14616
+    },
+    {
+      "epoch": 0.7453713478924963,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7773,
+      "step": 14617
+    },
+    {
+      "epoch": 0.7454223413486016,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 14618
+    },
+    {
+      "epoch": 0.745473334804707,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 14619
+    },
+    {
+      "epoch": 0.7455243282608124,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 14620
+    },
+    {
+      "epoch": 0.7455753217169178,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 14621
+    },
+    {
+      "epoch": 0.7456263151730231,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 14622
+    },
+    {
+      "epoch": 0.7456773086291286,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 14623
+    },
+    {
+      "epoch": 0.745728302085234,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8048,
+      "step": 14624
+    },
+    {
+      "epoch": 0.7457792955413394,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.002,
+      "loss": 2.8313,
+      "step": 14625
+    },
+    {
+      "epoch": 0.7458302889974447,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 14626
+    },
+    {
+      "epoch": 0.7458812824535501,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 14627
+    },
+    {
+      "epoch": 0.7459322759096555,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 14628
+    },
+    {
+      "epoch": 0.7459832693657609,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 14629
+    },
+    {
+      "epoch": 0.7460342628218664,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7460852562779717,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7971,
+      "step": 14631
+    },
+    {
+      "epoch": 0.7461362497340771,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 14632
+    },
+    {
+      "epoch": 0.7461872431901825,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 14633
+    },
+    {
+      "epoch": 0.7462382366462879,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 14634
+    },
+    {
+      "epoch": 0.7462892301023932,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 14635
+    },
+    {
+      "epoch": 0.7463402235584986,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 14636
+    },
+    {
+      "epoch": 0.7463912170146041,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 14637
+    },
+    {
+      "epoch": 0.7464422104707095,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 14638
+    },
+    {
+      "epoch": 0.7464932039268148,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 14639
+    },
+    {
+      "epoch": 0.7465441973829202,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7465951908390256,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 14641
+    },
+    {
+      "epoch": 0.746646184295131,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 14642
+    },
+    {
+      "epoch": 0.7466971777512363,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 14643
+    },
+    {
+      "epoch": 0.7467481712073418,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14644
+    },
+    {
+      "epoch": 0.7467991646634472,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 14645
+    },
+    {
+      "epoch": 0.7468501581195526,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 14646
+    },
+    {
+      "epoch": 0.746901151575658,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8318,
+      "step": 14647
+    },
+    {
+      "epoch": 0.7469521450317633,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 14648
+    },
+    {
+      "epoch": 0.7470031384878687,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14649
+    },
+    {
+      "epoch": 0.7470541319439741,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 14650
+    },
+    {
+      "epoch": 0.7471051254000796,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14651
+    },
+    {
+      "epoch": 0.7471561188561849,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 14652
+    },
+    {
+      "epoch": 0.7472071123122903,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 14653
+    },
+    {
+      "epoch": 0.7472581057683957,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 14654
+    },
+    {
+      "epoch": 0.7473090992245011,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14655
+    },
+    {
+      "epoch": 0.7473600926806064,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 14656
+    },
+    {
+      "epoch": 0.7474110861367118,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 14657
+    },
+    {
+      "epoch": 0.7474620795928173,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 14658
+    },
+    {
+      "epoch": 0.7475130730489227,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7901,
+      "step": 14659
+    },
+    {
+      "epoch": 0.747564066505028,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 14660
+    },
+    {
+      "epoch": 0.7476150599611334,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 14661
+    },
+    {
+      "epoch": 0.7476660534172388,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 14662
+    },
+    {
+      "epoch": 0.7477170468733442,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.7836,
+      "step": 14663
+    },
+    {
+      "epoch": 0.7477680403294495,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 14664
+    },
+    {
+      "epoch": 0.747819033785555,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8367,
+      "step": 14665
+    },
+    {
+      "epoch": 0.7478700272416604,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 14666
+    },
+    {
+      "epoch": 0.7479210206977658,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7919,
+      "step": 14667
+    },
+    {
+      "epoch": 0.7479720141538712,
+      "grad_norm": 2.0,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 14668
+    },
+    {
+      "epoch": 0.7480230076099765,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7879,
+      "step": 14669
+    },
+    {
+      "epoch": 0.7480740010660819,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7481249945221873,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 14671
+    },
+    {
+      "epoch": 0.7481759879782928,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 14672
+    },
+    {
+      "epoch": 0.7482269814343981,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8139,
+      "step": 14673
+    },
+    {
+      "epoch": 0.7482779748905035,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 14674
+    },
+    {
+      "epoch": 0.7483289683466089,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 14675
+    },
+    {
+      "epoch": 0.7483799618027143,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8188,
+      "step": 14676
+    },
+    {
+      "epoch": 0.7484309552588196,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 14677
+    },
+    {
+      "epoch": 0.748481948714925,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 14678
+    },
+    {
+      "epoch": 0.7485329421710305,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 14679
+    },
+    {
+      "epoch": 0.7485839356271359,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7486349290832413,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 14681
+    },
+    {
+      "epoch": 0.7486859225393466,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 14682
+    },
+    {
+      "epoch": 0.748736915995452,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 14683
+    },
+    {
+      "epoch": 0.7487879094515574,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 14684
+    },
+    {
+      "epoch": 0.7488389029076628,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 14685
+    },
+    {
+      "epoch": 0.7488898963637682,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 14686
+    },
+    {
+      "epoch": 0.7489408898198736,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8337,
+      "step": 14687
+    },
+    {
+      "epoch": 0.748991883275979,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 14688
+    },
+    {
+      "epoch": 0.7490428767320844,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14689
+    },
+    {
+      "epoch": 0.7490938701881897,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 14690
+    },
+    {
+      "epoch": 0.7491448636442951,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7923,
+      "step": 14691
+    },
+    {
+      "epoch": 0.7491958571004005,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 14692
+    },
+    {
+      "epoch": 0.749246850556506,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 14693
+    },
+    {
+      "epoch": 0.7492978440126113,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14694
+    },
+    {
+      "epoch": 0.7493488374687167,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 14695
+    },
+    {
+      "epoch": 0.7493998309248221,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14696
+    },
+    {
+      "epoch": 0.7494508243809275,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 14697
+    },
+    {
+      "epoch": 0.7495018178370328,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 14698
+    },
+    {
+      "epoch": 0.7495528112931382,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 14699
+    },
+    {
+      "epoch": 0.7496038047492437,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7496547982053491,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 14701
+    },
+    {
+      "epoch": 0.7497057916614545,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 14702
+    },
+    {
+      "epoch": 0.7497567851175598,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 14703
+    },
+    {
+      "epoch": 0.7498077785736652,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 14704
+    },
+    {
+      "epoch": 0.7498587720297706,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 14705
+    },
+    {
+      "epoch": 0.749909765485876,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 14706
+    },
+    {
+      "epoch": 0.7499607589419814,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8224,
+      "step": 14707
+    },
+    {
+      "epoch": 0.7500117523980868,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 14708
+    },
+    {
+      "epoch": 0.7500627458541922,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 14709
+    },
+    {
+      "epoch": 0.7501137393102976,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7501647327664029,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 14711
+    },
+    {
+      "epoch": 0.7502157262225083,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 14712
+    },
+    {
+      "epoch": 0.7502667196786137,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 14713
+    },
+    {
+      "epoch": 0.7503177131347192,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 14714
+    },
+    {
+      "epoch": 0.7503687065908246,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 14715
+    },
+    {
+      "epoch": 0.7504197000469299,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8184,
+      "step": 14716
+    },
+    {
+      "epoch": 0.7504706935030353,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 14717
+    },
+    {
+      "epoch": 0.7505216869591407,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8417,
+      "step": 14718
+    },
+    {
+      "epoch": 0.750572680415246,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 2.8326,
+      "step": 14719
+    },
+    {
+      "epoch": 0.7506236738713514,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 14720
+    },
+    {
+      "epoch": 0.7506746673274569,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 14721
+    },
+    {
+      "epoch": 0.7507256607835623,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 14722
+    },
+    {
+      "epoch": 0.7507766542396677,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8014,
+      "step": 14723
+    },
+    {
+      "epoch": 0.750827647695773,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 14724
+    },
+    {
+      "epoch": 0.7508786411518784,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 14725
+    },
+    {
+      "epoch": 0.7509296346079838,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 14726
+    },
+    {
+      "epoch": 0.7509806280640892,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14727
+    },
+    {
+      "epoch": 0.7510316215201946,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 14728
+    },
+    {
+      "epoch": 0.7510826149763,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 14729
+    },
+    {
+      "epoch": 0.7511336084324054,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7511846018885108,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 14731
+    },
+    {
+      "epoch": 0.7512355953446161,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 14732
+    },
+    {
+      "epoch": 0.7512865888007215,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 14733
+    },
+    {
+      "epoch": 0.7513375822568269,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 14734
+    },
+    {
+      "epoch": 0.7513885757129324,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 14735
+    },
+    {
+      "epoch": 0.7514395691690378,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 14736
+    },
+    {
+      "epoch": 0.7514905626251431,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8375,
+      "step": 14737
+    },
+    {
+      "epoch": 0.7515415560812485,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 14738
+    },
+    {
+      "epoch": 0.7515925495373539,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8056,
+      "step": 14739
+    },
+    {
+      "epoch": 0.7516435429934593,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8179,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7516945364495646,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 14741
+    },
+    {
+      "epoch": 0.7517455299056701,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 14742
+    },
+    {
+      "epoch": 0.7517965233617755,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 14743
+    },
+    {
+      "epoch": 0.7518475168178809,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 14744
+    },
+    {
+      "epoch": 0.7518985102739862,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14745
+    },
+    {
+      "epoch": 0.7519495037300916,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7844,
+      "step": 14746
+    },
+    {
+      "epoch": 0.752000497186197,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 14747
+    },
+    {
+      "epoch": 0.7520514906423024,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8067,
+      "step": 14748
+    },
+    {
+      "epoch": 0.7521024840984079,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 14749
+    },
+    {
+      "epoch": 0.7521534775545132,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7522044710106186,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 14751
+    },
+    {
+      "epoch": 0.752255464466724,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 14752
+    },
+    {
+      "epoch": 0.7523064579228294,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 14753
+    },
+    {
+      "epoch": 0.7523574513789347,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 14754
+    },
+    {
+      "epoch": 0.7524084448350401,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 14755
+    },
+    {
+      "epoch": 0.7524594382911456,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 14756
+    },
+    {
+      "epoch": 0.752510431747251,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.824,
+      "step": 14757
+    },
+    {
+      "epoch": 0.7525614252033563,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.819,
+      "step": 14758
+    },
+    {
+      "epoch": 0.7526124186594617,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 14759
+    },
+    {
+      "epoch": 0.7526634121155671,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7527144055716725,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 14761
+    },
+    {
+      "epoch": 0.7527653990277778,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 14762
+    },
+    {
+      "epoch": 0.7528163924838833,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.798,
+      "step": 14763
+    },
+    {
+      "epoch": 0.7528673859399887,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 14764
+    },
+    {
+      "epoch": 0.7529183793960941,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 14765
+    },
+    {
+      "epoch": 0.7529693728521994,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 14766
+    },
+    {
+      "epoch": 0.7530203663083048,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7964,
+      "step": 14767
+    },
+    {
+      "epoch": 0.7530713597644102,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 14768
+    },
+    {
+      "epoch": 0.7531223532205156,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 14769
+    },
+    {
+      "epoch": 0.7531733466766211,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8172,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7532243401327264,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7768,
+      "step": 14771
+    },
+    {
+      "epoch": 0.7532753335888318,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14772
+    },
+    {
+      "epoch": 0.7533263270449372,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.829,
+      "step": 14773
+    },
+    {
+      "epoch": 0.7533773205010426,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 14774
+    },
+    {
+      "epoch": 0.7534283139571479,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 14775
+    },
+    {
+      "epoch": 0.7534793074132533,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 14776
+    },
+    {
+      "epoch": 0.7535303008693588,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 14777
+    },
+    {
+      "epoch": 0.7535812943254642,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 14778
+    },
+    {
+      "epoch": 0.7536322877815695,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 14779
+    },
+    {
+      "epoch": 0.7536832812376749,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7766,
+      "step": 14780
+    },
+    {
+      "epoch": 0.7537342746937803,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 14781
+    },
+    {
+      "epoch": 0.7537852681498857,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 14782
+    },
+    {
+      "epoch": 0.753836261605991,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 14783
+    },
+    {
+      "epoch": 0.7538872550620965,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 14784
+    },
+    {
+      "epoch": 0.7539382485182019,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 14785
+    },
+    {
+      "epoch": 0.7539892419743073,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 14786
+    },
+    {
+      "epoch": 0.7540402354304127,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 14787
+    },
+    {
+      "epoch": 0.754091228886518,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 14788
+    },
+    {
+      "epoch": 0.7541422223426234,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.827,
+      "step": 14789
+    },
+    {
+      "epoch": 0.7541932157987288,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7542442092548343,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 14791
+    },
+    {
+      "epoch": 0.7542952027109396,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14792
+    },
+    {
+      "epoch": 0.754346196167045,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8145,
+      "step": 14793
+    },
+    {
+      "epoch": 0.7543971896231504,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 14794
+    },
+    {
+      "epoch": 0.7544481830792558,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8204,
+      "step": 14795
+    },
+    {
+      "epoch": 0.7544991765353611,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7671,
+      "step": 14796
+    },
+    {
+      "epoch": 0.7545501699914665,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8117,
+      "step": 14797
+    },
+    {
+      "epoch": 0.754601163447572,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8439,
+      "step": 14798
+    },
+    {
+      "epoch": 0.7546521569036774,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 14799
+    },
+    {
+      "epoch": 0.7547031503597827,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7547541438158881,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 14801
+    },
+    {
+      "epoch": 0.7548051372719935,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 14802
+    },
+    {
+      "epoch": 0.7548561307280989,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 14803
+    },
+    {
+      "epoch": 0.7549071241842042,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 14804
+    },
+    {
+      "epoch": 0.7549581176403097,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14805
+    },
+    {
+      "epoch": 0.7550091110964151,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 14806
+    },
+    {
+      "epoch": 0.7550601045525205,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14807
+    },
+    {
+      "epoch": 0.7551110980086259,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 14808
+    },
+    {
+      "epoch": 0.7551620914647312,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 14809
+    },
+    {
+      "epoch": 0.7552130849208366,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7944,
+      "step": 14810
+    },
+    {
+      "epoch": 0.755264078376942,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 14811
+    },
+    {
+      "epoch": 0.7553150718330475,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8227,
+      "step": 14812
+    },
+    {
+      "epoch": 0.7553660652891528,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 14813
+    },
+    {
+      "epoch": 0.7554170587452582,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 14814
+    },
+    {
+      "epoch": 0.7554680522013636,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 14815
+    },
+    {
+      "epoch": 0.755519045657469,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 14816
+    },
+    {
+      "epoch": 0.7555700391135743,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 14817
+    },
+    {
+      "epoch": 0.7556210325696797,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7756,
+      "step": 14818
+    },
+    {
+      "epoch": 0.7556720260257852,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 14819
+    },
+    {
+      "epoch": 0.7557230194818906,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8186,
+      "step": 14820
+    },
+    {
+      "epoch": 0.755774012937996,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 14821
+    },
+    {
+      "epoch": 0.7558250063941013,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 14822
+    },
+    {
+      "epoch": 0.7558759998502067,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 14823
+    },
+    {
+      "epoch": 0.7559269933063121,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7828,
+      "step": 14824
+    },
+    {
+      "epoch": 0.7559779867624175,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 14825
+    },
+    {
+      "epoch": 0.7560289802185229,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8556,
+      "step": 14826
+    },
+    {
+      "epoch": 0.7560799736746283,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 14827
+    },
+    {
+      "epoch": 0.7561309671307337,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 14828
+    },
+    {
+      "epoch": 0.7561819605868391,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 14829
+    },
+    {
+      "epoch": 0.7562329540429444,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7962,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7562839474990498,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8006,
+      "step": 14831
+    },
+    {
+      "epoch": 0.7563349409551552,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7761,
+      "step": 14832
+    },
+    {
+      "epoch": 0.7563859344112607,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7905,
+      "step": 14833
+    },
+    {
+      "epoch": 0.756436927867366,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.785,
+      "step": 14834
+    },
+    {
+      "epoch": 0.7564879213234714,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 14835
+    },
+    {
+      "epoch": 0.7565389147795768,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8273,
+      "step": 14836
+    },
+    {
+      "epoch": 0.7565899082356822,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 14837
+    },
+    {
+      "epoch": 0.7566409016917875,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.786,
+      "step": 14838
+    },
+    {
+      "epoch": 0.7566918951478929,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 14839
+    },
+    {
+      "epoch": 0.7567428886039984,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7567938820601038,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 14841
+    },
+    {
+      "epoch": 0.7568448755162092,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7908,
+      "step": 14842
+    },
+    {
+      "epoch": 0.7568958689723145,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 14843
+    },
+    {
+      "epoch": 0.7569468624284199,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 14844
+    },
+    {
+      "epoch": 0.7569978558845253,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 14845
+    },
+    {
+      "epoch": 0.7570488493406307,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7777,
+      "step": 14846
+    },
+    {
+      "epoch": 0.7570998427967361,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8108,
+      "step": 14847
+    },
+    {
+      "epoch": 0.7571508362528415,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7869,
+      "step": 14848
+    },
+    {
+      "epoch": 0.7572018297089469,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7969,
+      "step": 14849
+    },
+    {
+      "epoch": 0.7572528231650523,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7573038166211576,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 14851
+    },
+    {
+      "epoch": 0.757354810077263,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 14852
+    },
+    {
+      "epoch": 0.7574058035333684,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 14853
+    },
+    {
+      "epoch": 0.7574567969894739,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 14854
+    },
+    {
+      "epoch": 0.7575077904455793,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7929,
+      "step": 14855
+    },
+    {
+      "epoch": 0.7575587839016846,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8389,
+      "step": 14856
+    },
+    {
+      "epoch": 0.75760977735779,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7961,
+      "step": 14857
+    },
+    {
+      "epoch": 0.7576607708138954,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 14858
+    },
+    {
+      "epoch": 0.7577117642700008,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 14859
+    },
+    {
+      "epoch": 0.7577627577261061,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8091,
+      "step": 14860
+    },
+    {
+      "epoch": 0.7578137511822116,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 14861
+    },
+    {
+      "epoch": 0.757864744638317,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 14862
+    },
+    {
+      "epoch": 0.7579157380944224,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8352,
+      "step": 14863
+    },
+    {
+      "epoch": 0.7579667315505277,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 14864
+    },
+    {
+      "epoch": 0.7580177250066331,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 14865
+    },
+    {
+      "epoch": 0.7580687184627385,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 14866
+    },
+    {
+      "epoch": 0.7581197119188439,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 14867
+    },
+    {
+      "epoch": 0.7581707053749493,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8348,
+      "step": 14868
+    },
+    {
+      "epoch": 0.7582216988310547,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.832,
+      "step": 14869
+    },
+    {
+      "epoch": 0.7582726922871601,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7583236857432655,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8126,
+      "step": 14871
+    },
+    {
+      "epoch": 0.7583746791993708,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 14872
+    },
+    {
+      "epoch": 0.7584256726554762,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14873
+    },
+    {
+      "epoch": 0.7584766661115816,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 14874
+    },
+    {
+      "epoch": 0.7585276595676871,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 14875
+    },
+    {
+      "epoch": 0.7585786530237925,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8023,
+      "step": 14876
+    },
+    {
+      "epoch": 0.7586296464798978,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 14877
+    },
+    {
+      "epoch": 0.7586806399360032,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 14878
+    },
+    {
+      "epoch": 0.7587316333921086,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8397,
+      "step": 14879
+    },
+    {
+      "epoch": 0.758782626848214,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7588336203043193,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 14881
+    },
+    {
+      "epoch": 0.7588846137604248,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 14882
+    },
+    {
+      "epoch": 0.7589356072165302,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8414,
+      "step": 14883
+    },
+    {
+      "epoch": 0.7589866006726356,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 14884
+    },
+    {
+      "epoch": 0.7590375941287409,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 14885
+    },
+    {
+      "epoch": 0.7590885875848463,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7937,
+      "step": 14886
+    },
+    {
+      "epoch": 0.7591395810409517,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 14887
+    },
+    {
+      "epoch": 0.7591905744970571,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 14888
+    },
+    {
+      "epoch": 0.7592415679531626,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 14889
+    },
+    {
+      "epoch": 0.7592925614092679,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 14890
+    },
+    {
+      "epoch": 0.7593435548653733,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 14891
+    },
+    {
+      "epoch": 0.7593945483214787,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 14892
+    },
+    {
+      "epoch": 0.759445541777584,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8268,
+      "step": 14893
+    },
+    {
+      "epoch": 0.7594965352336894,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 14894
+    },
+    {
+      "epoch": 0.7595475286897948,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 14895
+    },
+    {
+      "epoch": 0.7595985221459003,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 14896
+    },
+    {
+      "epoch": 0.7596495156020057,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 14897
+    },
+    {
+      "epoch": 0.759700509058111,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14898
+    },
+    {
+      "epoch": 0.7597515025142164,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8169,
+      "step": 14899
+    },
+    {
+      "epoch": 0.7598024959703218,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7598534894264272,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7631,
+      "step": 14901
+    },
+    {
+      "epoch": 0.7599044828825325,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 14902
+    },
+    {
+      "epoch": 0.759955476338638,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 14903
+    },
+    {
+      "epoch": 0.7600064697947434,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 14904
+    },
+    {
+      "epoch": 0.7600574632508488,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8119,
+      "step": 14905
+    },
+    {
+      "epoch": 0.7601084567069541,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7941,
+      "step": 14906
+    },
+    {
+      "epoch": 0.7601594501630595,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 14907
+    },
+    {
+      "epoch": 0.7602104436191649,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 14908
+    },
+    {
+      "epoch": 0.7602614370752703,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7885,
+      "step": 14909
+    },
+    {
+      "epoch": 0.7603124305313758,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 14910
+    },
+    {
+      "epoch": 0.7603634239874811,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.7787,
+      "step": 14911
+    },
+    {
+      "epoch": 0.7604144174435865,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 14912
+    },
+    {
+      "epoch": 0.7604654108996919,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 14913
+    },
+    {
+      "epoch": 0.7605164043557973,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 14914
+    },
+    {
+      "epoch": 0.7605673978119026,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7818,
+      "step": 14915
+    },
+    {
+      "epoch": 0.760618391268008,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7857,
+      "step": 14916
+    },
+    {
+      "epoch": 0.7606693847241135,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8159,
+      "step": 14917
+    },
+    {
+      "epoch": 0.7607203781802189,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 14918
+    },
+    {
+      "epoch": 0.7607713716363242,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7801,
+      "step": 14919
+    },
+    {
+      "epoch": 0.7608223650924296,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 14920
+    },
+    {
+      "epoch": 0.760873358548535,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7887,
+      "step": 14921
+    },
+    {
+      "epoch": 0.7609243520046404,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14922
+    },
+    {
+      "epoch": 0.7609753454607457,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7763,
+      "step": 14923
+    },
+    {
+      "epoch": 0.7610263389168512,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 14924
+    },
+    {
+      "epoch": 0.7610773323729566,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 14925
+    },
+    {
+      "epoch": 0.761128325829062,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7981,
+      "step": 14926
+    },
+    {
+      "epoch": 0.7611793192851674,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 14927
+    },
+    {
+      "epoch": 0.7612303127412727,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 14928
+    },
+    {
+      "epoch": 0.7612813061973781,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7712,
+      "step": 14929
+    },
+    {
+      "epoch": 0.7613322996534835,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8154,
+      "step": 14930
+    },
+    {
+      "epoch": 0.761383293109589,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 14931
+    },
+    {
+      "epoch": 0.7614342865656943,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 14932
+    },
+    {
+      "epoch": 0.7614852800217997,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8096,
+      "step": 14933
+    },
+    {
+      "epoch": 0.7615362734779051,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 14934
+    },
+    {
+      "epoch": 0.7615872669340105,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.776,
+      "step": 14935
+    },
+    {
+      "epoch": 0.7616382603901158,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.8278,
+      "step": 14936
+    },
+    {
+      "epoch": 0.7616892538462212,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8001,
+      "step": 14937
+    },
+    {
+      "epoch": 0.7617402473023267,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 14938
+    },
+    {
+      "epoch": 0.7617912407584321,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8223,
+      "step": 14939
+    },
+    {
+      "epoch": 0.7618422342145375,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7618932276706428,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 14941
+    },
+    {
+      "epoch": 0.7619442211267482,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.811,
+      "step": 14942
+    },
+    {
+      "epoch": 0.7619952145828536,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 14943
+    },
+    {
+      "epoch": 0.762046208038959,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 14944
+    },
+    {
+      "epoch": 0.7620972014950644,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7803,
+      "step": 14945
+    },
+    {
+      "epoch": 0.7621481949511698,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 14946
+    },
+    {
+      "epoch": 0.7621991884072752,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.769,
+      "step": 14947
+    },
+    {
+      "epoch": 0.7622501818633806,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8011,
+      "step": 14948
+    },
+    {
+      "epoch": 0.7623011753194859,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8008,
+      "step": 14949
+    },
+    {
+      "epoch": 0.7623521687755913,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7624031622316967,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 14951
+    },
+    {
+      "epoch": 0.7624541556878022,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8161,
+      "step": 14952
+    },
+    {
+      "epoch": 0.7625051491439075,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8155,
+      "step": 14953
+    },
+    {
+      "epoch": 0.7625561426000129,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 14954
+    },
+    {
+      "epoch": 0.7626071360561183,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 14955
+    },
+    {
+      "epoch": 0.7626581295122237,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 14956
+    },
+    {
+      "epoch": 0.762709122968329,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8015,
+      "step": 14957
+    },
+    {
+      "epoch": 0.7627601164244344,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 14958
+    },
+    {
+      "epoch": 0.7628111098805399,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.7884,
+      "step": 14959
+    },
+    {
+      "epoch": 0.7628621033366453,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7629130967927507,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 14961
+    },
+    {
+      "epoch": 0.762964090248856,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.7953,
+      "step": 14962
+    },
+    {
+      "epoch": 0.7630150837049614,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 14963
+    },
+    {
+      "epoch": 0.7630660771610668,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 14964
+    },
+    {
+      "epoch": 0.7631170706171722,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 14965
+    },
+    {
+      "epoch": 0.7631680640732776,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14966
+    },
+    {
+      "epoch": 0.763219057529383,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 14967
+    },
+    {
+      "epoch": 0.7632700509854884,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 2.7983,
+      "step": 14968
+    },
+    {
+      "epoch": 0.7633210444415938,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 14969
+    },
+    {
+      "epoch": 0.7633720378976991,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.823,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7634230313538045,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.856,
+      "step": 14971
+    },
+    {
+      "epoch": 0.7634740248099099,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7998,
+      "step": 14972
+    },
+    {
+      "epoch": 0.7635250182660154,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 14973
+    },
+    {
+      "epoch": 0.7635760117221208,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 14974
+    },
+    {
+      "epoch": 0.7636270051782261,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 14975
+    },
+    {
+      "epoch": 0.7636779986343315,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 14976
+    },
+    {
+      "epoch": 0.7637289920904369,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.7939,
+      "step": 14977
+    },
+    {
+      "epoch": 0.7637799855465423,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8341,
+      "step": 14978
+    },
+    {
+      "epoch": 0.7638309790026476,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 14979
+    },
+    {
+      "epoch": 0.7638819724587531,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8135,
+      "step": 14980
+    },
+    {
+      "epoch": 0.7639329659148585,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.7952,
+      "step": 14981
+    },
+    {
+      "epoch": 0.7639839593709639,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 14982
+    },
+    {
+      "epoch": 0.7640349528270692,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 14983
+    },
+    {
+      "epoch": 0.7640859462831746,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 14984
+    },
+    {
+      "epoch": 0.76413693973928,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 14985
+    },
+    {
+      "epoch": 0.7641879331953854,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.788,
+      "step": 14986
+    },
+    {
+      "epoch": 0.7642389266514908,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 14987
+    },
+    {
+      "epoch": 0.7642899201075962,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.808,
+      "step": 14988
+    },
+    {
+      "epoch": 0.7643409135637016,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 14989
+    },
+    {
+      "epoch": 0.764391907019807,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7644429004759123,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 14991
+    },
+    {
+      "epoch": 0.7644938939320177,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 14992
+    },
+    {
+      "epoch": 0.7645448873881231,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8097,
+      "step": 14993
+    },
+    {
+      "epoch": 0.7645958808442286,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 14994
+    },
+    {
+      "epoch": 0.764646874300334,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 14995
+    },
+    {
+      "epoch": 0.7646978677564393,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 14996
+    },
+    {
+      "epoch": 0.7647488612125447,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 14997
+    },
+    {
+      "epoch": 0.7647998546686501,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.002,
+      "loss": 2.8241,
+      "step": 14998
+    },
+    {
+      "epoch": 0.7648508481247555,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 2.8051,
+      "step": 14999
+    },
+    {
+      "epoch": 0.7649018415808608,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8068,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7649528350369663,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 15001
+    },
+    {
+      "epoch": 0.7650038284930717,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 15002
+    },
+    {
+      "epoch": 0.7650548219491771,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.7946,
+      "step": 15003
+    },
+    {
+      "epoch": 0.7651058154052824,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 15004
+    },
+    {
+      "epoch": 0.7651568088613878,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 15005
+    },
+    {
+      "epoch": 0.7652078023174932,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 15006
+    },
+    {
+      "epoch": 0.7652587957735986,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8413,
+      "step": 15007
+    },
+    {
+      "epoch": 0.765309789229704,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 15008
+    },
+    {
+      "epoch": 0.7653607826858094,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.817,
+      "step": 15009
+    },
+    {
+      "epoch": 0.7654117761419148,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.792,
+      "step": 15010
+    },
+    {
+      "epoch": 0.7654627695980202,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 15011
+    },
+    {
+      "epoch": 0.7655137630541256,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 15012
+    },
+    {
+      "epoch": 0.7655647565102309,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 15013
+    },
+    {
+      "epoch": 0.7656157499663363,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 15014
+    },
+    {
+      "epoch": 0.7656667434224418,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.002,
+      "loss": 2.7833,
+      "step": 15015
+    },
+    {
+      "epoch": 0.7657177368785472,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 15016
+    },
+    {
+      "epoch": 0.7657687303346525,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7835,
+      "step": 15017
+    },
+    {
+      "epoch": 0.7658197237907579,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 15018
+    },
+    {
+      "epoch": 0.7658707172468633,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 15019
+    },
+    {
+      "epoch": 0.7659217107029687,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 15020
+    },
+    {
+      "epoch": 0.765972704159074,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 15021
+    },
+    {
+      "epoch": 0.7660236976151795,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 15022
+    },
+    {
+      "epoch": 0.7660746910712849,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 15023
+    },
+    {
+      "epoch": 0.7661256845273903,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8146,
+      "step": 15024
+    },
+    {
+      "epoch": 0.7661766779834956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 15025
+    },
+    {
+      "epoch": 0.766227671439601,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 15026
+    },
+    {
+      "epoch": 0.7662786648957064,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8274,
+      "step": 15027
+    },
+    {
+      "epoch": 0.7663296583518118,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8034,
+      "step": 15028
+    },
+    {
+      "epoch": 0.7663806518079173,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 15029
+    },
+    {
+      "epoch": 0.7664316452640226,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8254,
+      "step": 15030
+    },
+    {
+      "epoch": 0.766482638720128,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7935,
+      "step": 15031
+    },
+    {
+      "epoch": 0.7665336321762334,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.7985,
+      "step": 15032
+    },
+    {
+      "epoch": 0.7665846256323388,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 15033
+    },
+    {
+      "epoch": 0.7666356190884441,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7726,
+      "step": 15034
+    },
+    {
+      "epoch": 0.7666866125445495,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7915,
+      "step": 15035
+    },
+    {
+      "epoch": 0.766737606000655,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8226,
+      "step": 15036
+    },
+    {
+      "epoch": 0.7667885994567604,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.7954,
+      "step": 15037
+    },
+    {
+      "epoch": 0.7668395929128657,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 15038
+    },
+    {
+      "epoch": 0.7668905863689711,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7911,
+      "step": 15039
+    },
+    {
+      "epoch": 0.7669415798250765,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.844,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7669925732811819,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 15041
+    },
+    {
+      "epoch": 0.7670435667372872,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 15042
+    },
+    {
+      "epoch": 0.7670945601933927,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.7892,
+      "step": 15043
+    },
+    {
+      "epoch": 0.7671455536494981,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 15044
+    },
+    {
+      "epoch": 0.7671965471056035,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 15045
+    },
+    {
+      "epoch": 0.7672475405617089,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8531,
+      "step": 15046
+    },
+    {
+      "epoch": 0.7672985340178142,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.812,
+      "step": 15047
+    },
+    {
+      "epoch": 0.7673495274739196,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 15048
+    },
+    {
+      "epoch": 0.767400520930025,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 15049
+    },
+    {
+      "epoch": 0.7674515143861305,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7675025078422358,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 15051
+    },
+    {
+      "epoch": 0.7675535012983412,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 2.7827,
+      "step": 15052
+    },
+    {
+      "epoch": 0.7676044947544466,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.8261,
+      "step": 15053
+    },
+    {
+      "epoch": 0.767655488210552,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.7684,
+      "step": 15054
+    },
+    {
+      "epoch": 0.7677064816666573,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7927,
+      "step": 15055
+    },
+    {
+      "epoch": 0.7677574751227627,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8322,
+      "step": 15056
+    },
+    {
+      "epoch": 0.7678084685788682,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 15057
+    },
+    {
+      "epoch": 0.7678594620349736,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 15058
+    },
+    {
+      "epoch": 0.767910455491079,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 15059
+    },
+    {
+      "epoch": 0.7679614489471843,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 15060
+    },
+    {
+      "epoch": 0.7680124424032897,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8222,
+      "step": 15061
+    },
+    {
+      "epoch": 0.7680634358593951,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 15062
+    },
+    {
+      "epoch": 0.7681144293155004,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.7829,
+      "step": 15063
+    },
+    {
+      "epoch": 0.7681654227716059,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 15064
+    },
+    {
+      "epoch": 0.7682164162277113,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 15065
+    },
+    {
+      "epoch": 0.7682674096838167,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.804,
+      "step": 15066
+    },
+    {
+      "epoch": 0.7683184031399221,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7904,
+      "step": 15067
+    },
+    {
+      "epoch": 0.7683693965960274,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 15068
+    },
+    {
+      "epoch": 0.7684203900521328,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 15069
+    },
+    {
+      "epoch": 0.7684713835082382,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 15070
+    },
+    {
+      "epoch": 0.7685223769643437,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 15071
+    },
+    {
+      "epoch": 0.768573370420449,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 15072
+    },
+    {
+      "epoch": 0.7686243638765544,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 15073
+    },
+    {
+      "epoch": 0.7686753573326598,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 15074
+    },
+    {
+      "epoch": 0.7687263507887652,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8393,
+      "step": 15075
+    },
+    {
+      "epoch": 0.7687773442448705,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.8431,
+      "step": 15076
+    },
+    {
+      "epoch": 0.7688283377009759,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8238,
+      "step": 15077
+    },
+    {
+      "epoch": 0.7688793311570814,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 15078
+    },
+    {
+      "epoch": 0.7689303246131868,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8395,
+      "step": 15079
+    },
+    {
+      "epoch": 0.7689813180692922,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7690323115253975,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 15081
+    },
+    {
+      "epoch": 0.7690833049815029,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 15082
+    },
+    {
+      "epoch": 0.7691342984376083,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 15083
+    },
+    {
+      "epoch": 0.7691852918937137,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.7882,
+      "step": 15084
+    },
+    {
+      "epoch": 0.7692362853498191,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 15085
+    },
+    {
+      "epoch": 0.7692872788059245,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 15086
+    },
+    {
+      "epoch": 0.7693382722620299,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8147,
+      "step": 15087
+    },
+    {
+      "epoch": 0.7693892657181353,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8171,
+      "step": 15088
+    },
+    {
+      "epoch": 0.7694402591742406,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8115,
+      "step": 15089
+    },
+    {
+      "epoch": 0.769491252630346,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 15090
+    },
+    {
+      "epoch": 0.7695422460864514,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8055,
+      "step": 15091
+    },
+    {
+      "epoch": 0.7695932395425569,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8309,
+      "step": 15092
+    },
+    {
+      "epoch": 0.7696442329986622,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 15093
+    },
+    {
+      "epoch": 0.7696952264547676,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8198,
+      "step": 15094
+    },
+    {
+      "epoch": 0.769746219910873,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7947,
+      "step": 15095
+    },
+    {
+      "epoch": 0.7697972133669784,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8256,
+      "step": 15096
+    },
+    {
+      "epoch": 0.7698482068230837,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8044,
+      "step": 15097
+    },
+    {
+      "epoch": 0.7698992002791891,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 15098
+    },
+    {
+      "epoch": 0.7699501937352946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 2.8402,
+      "step": 15099
+    },
+    {
+      "epoch": 0.7700011871914,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7700521806475054,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 2.7824,
+      "step": 15101
+    },
+    {
+      "epoch": 0.7701031741036107,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8021,
+      "step": 15102
+    },
+    {
+      "epoch": 0.7701541675597161,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8122,
+      "step": 15103
+    },
+    {
+      "epoch": 0.7702051610158215,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 15104
+    },
+    {
+      "epoch": 0.7702561544719269,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 2.8396,
+      "step": 15105
+    },
+    {
+      "epoch": 0.7703071479280323,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 15106
+    },
+    {
+      "epoch": 0.7703581413841377,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 2.8144,
+      "step": 15107
+    },
+    {
+      "epoch": 0.7704091348402431,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8477,
+      "step": 15108
+    },
+    {
+      "epoch": 0.7704601282963485,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.8295,
+      "step": 15109
+    },
+    {
+      "epoch": 0.7705111217524538,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8218,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7705621152085592,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8374,
+      "step": 15111
+    },
+    {
+      "epoch": 0.7706131086646646,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8347,
+      "step": 15112
+    },
+    {
+      "epoch": 0.7706641021207701,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 15113
+    },
+    {
+      "epoch": 0.7707150955768755,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 15114
+    },
+    {
+      "epoch": 0.7707660890329808,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 15115
+    },
+    {
+      "epoch": 0.7708170824890862,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8164,
+      "step": 15116
+    },
+    {
+      "epoch": 0.7708680759451916,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 15117
+    },
+    {
+      "epoch": 0.770919069401297,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 15118
+    },
+    {
+      "epoch": 0.7709700628574023,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 15119
+    },
+    {
+      "epoch": 0.7710210563135078,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7710720497696132,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8077,
+      "step": 15121
+    },
+    {
+      "epoch": 0.7711230432257186,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7891,
+      "step": 15122
+    },
+    {
+      "epoch": 0.7711740366818239,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 15123
+    },
+    {
+      "epoch": 0.7712250301379293,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 15124
+    },
+    {
+      "epoch": 0.7712760235940347,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 15125
+    },
+    {
+      "epoch": 0.7713270170501401,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8069,
+      "step": 15126
+    },
+    {
+      "epoch": 0.7713780105062455,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 15127
+    },
+    {
+      "epoch": 0.7714290039623509,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 15128
+    },
+    {
+      "epoch": 0.7714799974184563,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 15129
+    },
+    {
+      "epoch": 0.7715309908745617,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 15130
+    },
+    {
+      "epoch": 0.771581984330667,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 15131
+    },
+    {
+      "epoch": 0.7716329777867724,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 15132
+    },
+    {
+      "epoch": 0.7716839712428778,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 15133
+    },
+    {
+      "epoch": 0.7717349646989833,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7982,
+      "step": 15134
+    },
+    {
+      "epoch": 0.7717859581550887,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 15135
+    },
+    {
+      "epoch": 0.771836951611194,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8279,
+      "step": 15136
+    },
+    {
+      "epoch": 0.7718879450672994,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.7865,
+      "step": 15137
+    },
+    {
+      "epoch": 0.7719389385234048,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 2.8332,
+      "step": 15138
+    },
+    {
+      "epoch": 0.7719899319795102,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 15139
+    },
+    {
+      "epoch": 0.7720409254356155,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8081,
+      "step": 15140
+    },
+    {
+      "epoch": 0.772091918891721,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7906,
+      "step": 15141
+    },
+    {
+      "epoch": 0.7721429123478264,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8085,
+      "step": 15142
+    },
+    {
+      "epoch": 0.7721939058039318,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 15143
+    },
+    {
+      "epoch": 0.7722448992600371,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 15144
+    },
+    {
+      "epoch": 0.7722958927161425,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.7967,
+      "step": 15145
+    },
+    {
+      "epoch": 0.7723468861722479,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 15146
+    },
+    {
+      "epoch": 0.7723978796283533,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 15147
+    },
+    {
+      "epoch": 0.7724488730844588,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 15148
+    },
+    {
+      "epoch": 0.7724998665405641,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8529,
+      "step": 15149
+    },
+    {
+      "epoch": 0.7725508599966695,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.7991,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7726018534527749,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.803,
+      "step": 15151
+    },
+    {
+      "epoch": 0.7726528469088803,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 15152
+    },
+    {
+      "epoch": 0.7727038403649856,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 15153
+    },
+    {
+      "epoch": 0.772754833821091,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 2.8176,
+      "step": 15154
+    },
+    {
+      "epoch": 0.7728058272771965,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8017,
+      "step": 15155
+    },
+    {
+      "epoch": 0.7728568207333019,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 2.8035,
+      "step": 15156
+    },
+    {
+      "epoch": 0.7729078141894072,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 15157
+    },
+    {
+      "epoch": 0.7729588076455126,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 15158
+    },
+    {
+      "epoch": 0.773009801101618,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8331,
+      "step": 15159
+    },
+    {
+      "epoch": 0.7730607945577234,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8094,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7731117880138287,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.002,
+      "loss": 2.7769,
+      "step": 15161
+    },
+    {
+      "epoch": 0.7731627814699342,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8316,
+      "step": 15162
+    },
+    {
+      "epoch": 0.7732137749260396,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8031,
+      "step": 15163
+    },
+    {
+      "epoch": 0.773264768382145,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 15164
+    },
+    {
+      "epoch": 0.7733157618382503,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 15165
+    },
+    {
+      "epoch": 0.7733667552943557,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 15166
+    },
+    {
+      "epoch": 0.7734177487504611,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 15167
+    },
+    {
+      "epoch": 0.7734687422065665,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 15168
+    },
+    {
+      "epoch": 0.773519735662672,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 15169
+    },
+    {
+      "epoch": 0.7735707291187773,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7909,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7736217225748827,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 15171
+    },
+    {
+      "epoch": 0.7736727160309881,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8233,
+      "step": 15172
+    },
+    {
+      "epoch": 0.7737237094870935,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8368,
+      "step": 15173
+    },
+    {
+      "epoch": 0.7737747029431988,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 15174
+    },
+    {
+      "epoch": 0.7738256963993042,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 15175
+    },
+    {
+      "epoch": 0.7738766898554097,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8221,
+      "step": 15176
+    },
+    {
+      "epoch": 0.7739276833115151,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 15177
+    },
+    {
+      "epoch": 0.7739786767676204,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8209,
+      "step": 15178
+    },
+    {
+      "epoch": 0.7740296702237258,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8,
+      "step": 15179
+    },
+    {
+      "epoch": 0.7740806636798312,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7993,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7741316571359366,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 2.8003,
+      "step": 15181
+    },
+    {
+      "epoch": 0.7741826505920419,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 15182
+    },
+    {
+      "epoch": 0.7742336440481474,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8327,
+      "step": 15183
+    },
+    {
+      "epoch": 0.7742846375042528,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8104,
+      "step": 15184
+    },
+    {
+      "epoch": 0.7743356309603582,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8336,
+      "step": 15185
+    },
+    {
+      "epoch": 0.7743866244164636,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 15186
+    },
+    {
+      "epoch": 0.7744376178725689,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 15187
+    },
+    {
+      "epoch": 0.7744886113286743,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8004,
+      "step": 15188
+    },
+    {
+      "epoch": 0.7745396047847797,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 15189
+    },
+    {
+      "epoch": 0.7745905982408852,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7746415916969905,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 2.8478,
+      "step": 15191
+    },
+    {
+      "epoch": 0.7746925851530959,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 15192
+    },
+    {
+      "epoch": 0.7747435786092013,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.842,
+      "step": 15193
+    },
+    {
+      "epoch": 0.7747945720653067,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8247,
+      "step": 15194
+    },
+    {
+      "epoch": 0.774845565521412,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.834,
+      "step": 15195
+    },
+    {
+      "epoch": 0.7748965589775174,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 15196
+    },
+    {
+      "epoch": 0.7749475524336229,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 15197
+    },
+    {
+      "epoch": 0.7749985458897283,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 2.7933,
+      "step": 15198
+    },
+    {
+      "epoch": 0.7750495393458336,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7818,
+      "step": 15199
+    },
+    {
+      "epoch": 0.775100532801939,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7975,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7751515262580444,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8253,
+      "step": 15201
+    },
+    {
+      "epoch": 0.7752025197141498,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 15202
+    },
+    {
+      "epoch": 0.7752535131702551,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 15203
+    },
+    {
+      "epoch": 0.7753045066263606,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7757,
+      "step": 15204
+    },
+    {
+      "epoch": 0.775355500082466,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8043,
+      "step": 15205
+    },
+    {
+      "epoch": 0.7754064935385714,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.7809,
+      "step": 15206
+    },
+    {
+      "epoch": 0.7754574869946768,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 15207
+    },
+    {
+      "epoch": 0.7755084804507821,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 15208
+    },
+    {
+      "epoch": 0.7755594739068875,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8343,
+      "step": 15209
+    },
+    {
+      "epoch": 0.7756104673629929,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8205,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7756614608190984,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 15211
+    },
+    {
+      "epoch": 0.7757124542752037,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8376,
+      "step": 15212
+    },
+    {
+      "epoch": 0.7757634477313091,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 15213
+    },
+    {
+      "epoch": 0.7758144411874145,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8438,
+      "step": 15214
+    },
+    {
+      "epoch": 0.7758654346435199,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 15215
+    },
+    {
+      "epoch": 0.7759164280996252,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 15216
+    },
+    {
+      "epoch": 0.7759674215557306,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7945,
+      "step": 15217
+    },
+    {
+      "epoch": 0.7760184150118361,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7986,
+      "step": 15218
+    },
+    {
+      "epoch": 0.7760694084679415,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8325,
+      "step": 15219
+    },
+    {
+      "epoch": 0.7761204019240469,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8076,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7761713953801522,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8027,
+      "step": 15221
+    },
+    {
+      "epoch": 0.7762223888362576,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.826,
+      "step": 15222
+    },
+    {
+      "epoch": 0.776273382292363,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8246,
+      "step": 15223
+    },
+    {
+      "epoch": 0.7763243757484684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 15224
+    },
+    {
+      "epoch": 0.7763753692045738,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 15225
+    },
+    {
+      "epoch": 0.7764263626606792,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8167,
+      "step": 15226
+    },
+    {
+      "epoch": 0.7764773561167846,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 2.8088,
+      "step": 15227
+    },
+    {
+      "epoch": 0.77652834957289,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 15228
+    },
+    {
+      "epoch": 0.7765793430289953,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 15229
+    },
+    {
+      "epoch": 0.7766303364851007,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7766813299412061,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 15231
+    },
+    {
+      "epoch": 0.7767323233973116,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 2.7809,
+      "step": 15232
+    },
+    {
+      "epoch": 0.776783316853417,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.822,
+      "step": 15233
+    },
+    {
+      "epoch": 0.7768343103095223,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 15234
+    },
+    {
+      "epoch": 0.7768853037656277,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 15235
+    },
+    {
+      "epoch": 0.7769362972217331,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8305,
+      "step": 15236
+    },
+    {
+      "epoch": 0.7769872906778384,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8219,
+      "step": 15237
+    },
+    {
+      "epoch": 0.7770382841339438,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.7907,
+      "step": 15238
+    },
+    {
+      "epoch": 0.7770892775900493,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7997,
+      "step": 15239
+    },
+    {
+      "epoch": 0.7771402710461547,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8284,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7771912645022601,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8314,
+      "step": 15241
+    },
+    {
+      "epoch": 0.7772422579583654,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 15242
+    },
+    {
+      "epoch": 0.7772932514144708,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 15243
+    },
+    {
+      "epoch": 0.7773442448705762,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8291,
+      "step": 15244
+    },
+    {
+      "epoch": 0.7773952383266816,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 15245
+    },
+    {
+      "epoch": 0.777446231782787,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.8185,
+      "step": 15246
+    },
+    {
+      "epoch": 0.7774972252388924,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.825,
+      "step": 15247
+    },
+    {
+      "epoch": 0.7775482186949978,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 15248
+    },
+    {
+      "epoch": 0.7775992121511032,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.813,
+      "step": 15249
+    },
+    {
+      "epoch": 0.7776502056072085,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8099,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7777011990633139,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 15251
+    },
+    {
+      "epoch": 0.7777521925194193,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7873,
+      "step": 15252
+    },
+    {
+      "epoch": 0.7778031859755248,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 15253
+    },
+    {
+      "epoch": 0.7778541794316302,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 15254
+    },
+    {
+      "epoch": 0.7779051728877355,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 15255
+    },
+    {
+      "epoch": 0.7779561663438409,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7894,
+      "step": 15256
+    },
+    {
+      "epoch": 0.7780071597999463,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8165,
+      "step": 15257
+    },
+    {
+      "epoch": 0.7780581532560517,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 15258
+    },
+    {
+      "epoch": 0.778109146712157,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 15259
+    },
+    {
+      "epoch": 0.7781601401682625,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7782111336243679,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 15261
+    },
+    {
+      "epoch": 0.7782621270804733,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8071,
+      "step": 15262
+    },
+    {
+      "epoch": 0.7783131205365786,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 15263
+    },
+    {
+      "epoch": 0.778364113992684,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8032,
+      "step": 15264
+    },
+    {
+      "epoch": 0.7784151074487894,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 2.8049,
+      "step": 15265
+    },
+    {
+      "epoch": 0.7784661009048948,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 15266
+    },
+    {
+      "epoch": 0.7785170943610003,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 15267
+    },
+    {
+      "epoch": 0.7785680878171056,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 15268
+    },
+    {
+      "epoch": 0.778619081273211,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 15269
+    },
+    {
+      "epoch": 0.7786700747293164,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 2.8083,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7787210681854217,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8217,
+      "step": 15271
+    },
+    {
+      "epoch": 0.7787720616415271,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.799,
+      "step": 15272
+    },
+    {
+      "epoch": 0.7788230550976325,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 15273
+    },
+    {
+      "epoch": 0.778874048553738,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 15274
+    },
+    {
+      "epoch": 0.7789250420098434,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 15275
+    },
+    {
+      "epoch": 0.7789760354659487,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7821,
+      "step": 15276
+    },
+    {
+      "epoch": 0.7790270289220541,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8257,
+      "step": 15277
+    },
+    {
+      "epoch": 0.7790780223781595,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.83,
+      "step": 15278
+    },
+    {
+      "epoch": 0.7791290158342649,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8116,
+      "step": 15279
+    },
+    {
+      "epoch": 0.7791800092903702,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.816,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7792310027464757,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8054,
+      "step": 15281
+    },
+    {
+      "epoch": 0.7792819962025811,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8028,
+      "step": 15282
+    },
+    {
+      "epoch": 0.7793329896586865,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8208,
+      "step": 15283
+    },
+    {
+      "epoch": 0.7793839831147918,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 2.8042,
+      "step": 15284
+    },
+    {
+      "epoch": 0.7794349765708972,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8175,
+      "step": 15285
+    },
+    {
+      "epoch": 0.7794859700270026,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.7783,
+      "step": 15286
+    },
+    {
+      "epoch": 0.779536963483108,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 15287
+    },
+    {
+      "epoch": 0.7795879569392135,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7778,
+      "step": 15288
+    },
+    {
+      "epoch": 0.7796389503953188,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 15289
+    },
+    {
+      "epoch": 0.7796899438514242,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7797409373075296,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 15291
+    },
+    {
+      "epoch": 0.779791930763635,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 15292
+    },
+    {
+      "epoch": 0.7798429242197403,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8062,
+      "step": 15293
+    },
+    {
+      "epoch": 0.7798939176758457,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8304,
+      "step": 15294
+    },
+    {
+      "epoch": 0.7799449111319512,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 15295
+    },
+    {
+      "epoch": 0.7799959045880566,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7897,
+      "step": 15296
+    },
+    {
+      "epoch": 0.7800468980441619,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 15297
+    },
+    {
+      "epoch": 0.7800978915002673,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.002,
+      "loss": 2.8349,
+      "step": 15298
+    },
+    {
+      "epoch": 0.7801488849563727,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 15299
+    },
+    {
+      "epoch": 0.7801998784124781,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7802508718685834,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8173,
+      "step": 15301
+    },
+    {
+      "epoch": 0.7803018653246889,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 15302
+    },
+    {
+      "epoch": 0.7803528587807943,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 15303
+    },
+    {
+      "epoch": 0.7804038522368997,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.7863,
+      "step": 15304
+    },
+    {
+      "epoch": 0.780454845693005,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 15305
+    },
+    {
+      "epoch": 0.7805058391491104,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8178,
+      "step": 15306
+    },
+    {
+      "epoch": 0.7805568326052158,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 15307
+    },
+    {
+      "epoch": 0.7806078260613212,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 15308
+    },
+    {
+      "epoch": 0.7806588195174267,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.7951,
+      "step": 15309
+    },
+    {
+      "epoch": 0.780709812973532,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.7838,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7807608064296374,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8148,
+      "step": 15311
+    },
+    {
+      "epoch": 0.7808117998857428,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.7837,
+      "step": 15312
+    },
+    {
+      "epoch": 0.7808627933418482,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 2.8163,
+      "step": 15313
+    },
+    {
+      "epoch": 0.7809137867979535,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8357,
+      "step": 15314
+    },
+    {
+      "epoch": 0.7809647802540589,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 15315
+    },
+    {
+      "epoch": 0.7810157737101644,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8133,
+      "step": 15316
+    },
+    {
+      "epoch": 0.7810667671662698,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 15317
+    },
+    {
+      "epoch": 0.7811177606223751,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 15318
+    },
+    {
+      "epoch": 0.7811687540784805,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 15319
+    },
+    {
+      "epoch": 0.7812197475345859,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7812707409906913,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 15321
+    },
+    {
+      "epoch": 0.7813217344467966,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.8012,
+      "step": 15322
+    },
+    {
+      "epoch": 0.7813727279029021,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8125,
+      "step": 15323
+    },
+    {
+      "epoch": 0.7814237213590075,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 15324
+    },
+    {
+      "epoch": 0.7814747148151129,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 15325
+    },
+    {
+      "epoch": 0.7815257082712183,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8075,
+      "step": 15326
+    },
+    {
+      "epoch": 0.7815767017273236,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8112,
+      "step": 15327
+    },
+    {
+      "epoch": 0.781627695183429,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8289,
+      "step": 15328
+    },
+    {
+      "epoch": 0.7816786886395344,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 2.8089,
+      "step": 15329
+    },
+    {
+      "epoch": 0.7817296820956399,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 2.8063,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7817806755517452,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8142,
+      "step": 15331
+    },
+    {
+      "epoch": 0.7818316690078506,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 2.7976,
+      "step": 15332
+    },
+    {
+      "epoch": 0.781882662463956,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 2.8152,
+      "step": 15333
+    },
+    {
+      "epoch": 0.7819336559200614,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.8388,
+      "step": 15334
+    },
+    {
+      "epoch": 0.7819846493761667,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8128,
+      "step": 15335
+    },
+    {
+      "epoch": 0.7820356428322721,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 15336
+    },
+    {
+      "epoch": 0.7820866362883776,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 15337
+    },
+    {
+      "epoch": 0.782137629744483,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8007,
+      "step": 15338
+    },
+    {
+      "epoch": 0.7821886232005884,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8158,
+      "step": 15339
+    },
+    {
+      "epoch": 0.7822396166566937,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7822906101127991,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8243,
+      "step": 15341
+    },
+    {
+      "epoch": 0.7823416035689045,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 15342
+    },
+    {
+      "epoch": 0.7823925970250099,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 15343
+    },
+    {
+      "epoch": 0.7824435904811153,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7926,
+      "step": 15344
+    },
+    {
+      "epoch": 0.7824945839372207,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 15345
+    },
+    {
+      "epoch": 0.7825455773933261,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.828,
+      "step": 15346
+    },
+    {
+      "epoch": 0.7825965708494315,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7972,
+      "step": 15347
+    },
+    {
+      "epoch": 0.7826475643055368,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8177,
+      "step": 15348
+    },
+    {
+      "epoch": 0.7826985577616422,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 15349
+    },
+    {
+      "epoch": 0.7827495512177476,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7828005446738531,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.779,
+      "step": 15351
+    },
+    {
+      "epoch": 0.7828515381299584,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7995,
+      "step": 15352
+    },
+    {
+      "epoch": 0.7829025315860638,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8181,
+      "step": 15353
+    },
+    {
+      "epoch": 0.7829535250421692,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8312,
+      "step": 15354
+    },
+    {
+      "epoch": 0.7830045184982746,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8153,
+      "step": 15355
+    },
+    {
+      "epoch": 0.78305551195438,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8013,
+      "step": 15356
+    },
+    {
+      "epoch": 0.7831065054104853,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8005,
+      "step": 15357
+    },
+    {
+      "epoch": 0.7831574988665908,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 2.7791,
+      "step": 15358
+    },
+    {
+      "epoch": 0.7832084923226962,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.7973,
+      "step": 15359
+    },
+    {
+      "epoch": 0.7832594857788016,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7886,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7833104792349069,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 15361
+    },
+    {
+      "epoch": 0.7833614726910123,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 15362
+    },
+    {
+      "epoch": 0.7834124661471177,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8019,
+      "step": 15363
+    },
+    {
+      "epoch": 0.7834634596032231,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8505,
+      "step": 15364
+    },
+    {
+      "epoch": 0.7835144530593285,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 2.809,
+      "step": 15365
+    },
+    {
+      "epoch": 0.7835654465154339,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 15366
+    },
+    {
+      "epoch": 0.7836164399715393,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.8512,
+      "step": 15367
+    },
+    {
+      "epoch": 0.7836674334276447,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 15368
+    },
+    {
+      "epoch": 0.78371842688375,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 15369
+    },
+    {
+      "epoch": 0.7837694203398554,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7838204137959608,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 15371
+    },
+    {
+      "epoch": 0.7838714072520663,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7896,
+      "step": 15372
+    },
+    {
+      "epoch": 0.7839224007081717,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8082,
+      "step": 15373
+    },
+    {
+      "epoch": 0.783973394164277,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 15374
+    },
+    {
+      "epoch": 0.7840243876203824,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 15375
+    },
+    {
+      "epoch": 0.7840753810764878,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 15376
+    },
+    {
+      "epoch": 0.7841263745325932,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 15377
+    },
+    {
+      "epoch": 0.7841773679886985,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8024,
+      "step": 15378
+    },
+    {
+      "epoch": 0.784228361444804,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8302,
+      "step": 15379
+    },
+    {
+      "epoch": 0.7842793549009094,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 2.8086,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7843303483570148,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7791,
+      "step": 15381
+    },
+    {
+      "epoch": 0.7843813418131201,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.8232,
+      "step": 15382
+    },
+    {
+      "epoch": 0.7844323352692255,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7924,
+      "step": 15383
+    },
+    {
+      "epoch": 0.7844833287253309,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7868,
+      "step": 15384
+    },
+    {
+      "epoch": 0.7845343221814363,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 15385
+    },
+    {
+      "epoch": 0.7845853156375417,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8047,
+      "step": 15386
+    },
+    {
+      "epoch": 0.7846363090936471,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8315,
+      "step": 15387
+    },
+    {
+      "epoch": 0.7846873025497525,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8371,
+      "step": 15388
+    },
+    {
+      "epoch": 0.7847382960058579,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 15389
+    },
+    {
+      "epoch": 0.7847892894619632,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 2.8321,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7848402829180686,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8275,
+      "step": 15391
+    },
+    {
+      "epoch": 0.784891276374174,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 15392
+    },
+    {
+      "epoch": 0.7849422698302795,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8009,
+      "step": 15393
+    },
+    {
+      "epoch": 0.7849932632863849,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 15394
+    },
+    {
+      "epoch": 0.7850442567424902,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 15395
+    },
+    {
+      "epoch": 0.7850952501985956,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8149,
+      "step": 15396
+    },
+    {
+      "epoch": 0.785146243654701,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 2.8093,
+      "step": 15397
+    },
+    {
+      "epoch": 0.7851972371108064,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.7992,
+      "step": 15398
+    },
+    {
+      "epoch": 0.7852482305669117,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 2.8504,
+      "step": 15399
+    },
+    {
+      "epoch": 0.7852992240230172,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 2.8287,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7853502174791226,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8237,
+      "step": 15401
+    },
+    {
+      "epoch": 0.785401210935228,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 15402
+    },
+    {
+      "epoch": 0.7854522043913333,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8379,
+      "step": 15403
+    },
+    {
+      "epoch": 0.7855031978474387,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.815,
+      "step": 15404
+    },
+    {
+      "epoch": 0.7855541913035441,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8382,
+      "step": 15405
+    },
+    {
+      "epoch": 0.7856051847596495,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7913,
+      "step": 15406
+    },
+    {
+      "epoch": 0.785656178215755,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 15407
+    },
+    {
+      "epoch": 0.7857071716718603,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 15408
+    },
+    {
+      "epoch": 0.7857581651279657,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7955,
+      "step": 15409
+    },
+    {
+      "epoch": 0.7858091585840711,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7858601520401765,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.7867,
+      "step": 15411
+    },
+    {
+      "epoch": 0.7859111454962818,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 15412
+    },
+    {
+      "epoch": 0.7859621389523872,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8029,
+      "step": 15413
+    },
+    {
+      "epoch": 0.7860131324084927,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 15414
+    },
+    {
+      "epoch": 0.7860641258645981,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7898,
+      "step": 15415
+    },
+    {
+      "epoch": 0.7861151193207034,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 15416
+    },
+    {
+      "epoch": 0.7861661127768088,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8255,
+      "step": 15417
+    },
+    {
+      "epoch": 0.7862171062329142,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 2.7978,
+      "step": 15418
+    },
+    {
+      "epoch": 0.7862680996890196,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8526,
+      "step": 15419
+    },
+    {
+      "epoch": 0.7863190931451249,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.8285,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7863700866012304,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8324,
+      "step": 15421
+    },
+    {
+      "epoch": 0.7864210800573358,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8239,
+      "step": 15422
+    },
+    {
+      "epoch": 0.7864720735134412,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.839,
+      "step": 15423
+    },
+    {
+      "epoch": 0.7865230669695465,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 15424
+    },
+    {
+      "epoch": 0.7865740604256519,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8419,
+      "step": 15425
+    },
+    {
+      "epoch": 0.7866250538817573,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 2.8045,
+      "step": 15426
+    },
+    {
+      "epoch": 0.7866760473378627,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7689,
+      "step": 15427
+    },
+    {
+      "epoch": 0.7867270407939682,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8058,
+      "step": 15428
+    },
+    {
+      "epoch": 0.7867780342500735,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.7765,
+      "step": 15429
+    },
+    {
+      "epoch": 0.7868290277061789,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8259,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7868800211622843,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 15431
+    },
+    {
+      "epoch": 0.7869310146183897,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8059,
+      "step": 15432
+    },
+    {
+      "epoch": 0.786982008074495,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 2.8514,
+      "step": 15433
+    },
+    {
+      "epoch": 0.7870330015306004,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7845,
+      "step": 15434
+    },
+    {
+      "epoch": 0.7870839949867059,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 15435
+    },
+    {
+      "epoch": 0.7871349884428113,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8303,
+      "step": 15436
+    },
+    {
+      "epoch": 0.7871859818989166,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 15437
+    },
+    {
+      "epoch": 0.787236975355022,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.7958,
+      "step": 15438
+    },
+    {
+      "epoch": 0.7872879688111274,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8174,
+      "step": 15439
+    },
+    {
+      "epoch": 0.7873389622672328,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7999,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7873899557233381,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8301,
+      "step": 15441
+    },
+    {
+      "epoch": 0.7874409491794436,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8132,
+      "step": 15442
+    },
+    {
+      "epoch": 0.787491942635549,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.748,
+      "step": 15443
+    },
+    {
+      "epoch": 0.7875429360916544,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.7634,
+      "step": 15444
+    },
+    {
+      "epoch": 0.7875939295477598,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8235,
+      "step": 15445
+    },
+    {
+      "epoch": 0.7876449230038651,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8306,
+      "step": 15446
+    },
+    {
+      "epoch": 0.7876959164599705,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 15447
+    },
+    {
+      "epoch": 0.7877469099160759,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 2.8061,
+      "step": 15448
+    },
+    {
+      "epoch": 0.7877979033721814,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 2.8328,
+      "step": 15449
+    },
+    {
+      "epoch": 0.7878488968282867,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.7916,
+      "step": 15450
+    },
+    {
+      "epoch": 0.7878998902843921,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8138,
+      "step": 15451
+    },
+    {
+      "epoch": 0.7879508837404975,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8206,
+      "step": 15452
+    },
+    {
+      "epoch": 0.7880018771966029,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.7936,
+      "step": 15453
+    },
+    {
+      "epoch": 0.7880528706527082,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.8201,
+      "step": 15454
+    },
+    {
+      "epoch": 0.7881038641088136,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 15455
+    },
+    {
+      "epoch": 0.7881548575649191,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8299,
+      "step": 15456
+    },
+    {
+      "epoch": 0.7882058510210245,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8065,
+      "step": 15457
+    },
+    {
+      "epoch": 0.7882568444771298,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8189,
+      "step": 15458
+    },
+    {
+      "epoch": 0.7883078379332352,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 15459
+    },
+    {
+      "epoch": 0.7883588313893406,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.806,
+      "step": 15460
+    },
+    {
+      "epoch": 0.788409824845446,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.7795,
+      "step": 15461
+    },
+    {
+      "epoch": 0.7884608183015513,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 15462
+    },
+    {
+      "epoch": 0.7885118117576568,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.807,
+      "step": 15463
+    },
+    {
+      "epoch": 0.7885628052137622,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.002,
+      "loss": 2.82,
+      "step": 15464
+    },
+    {
+      "epoch": 0.7886137986698676,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8123,
+      "step": 15465
+    },
+    {
+      "epoch": 0.788664792125973,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 15466
+    },
+    {
+      "epoch": 0.7887157855820783,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8317,
+      "step": 15467
+    },
+    {
+      "epoch": 0.7887667790381837,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.7918,
+      "step": 15468
+    },
+    {
+      "epoch": 0.7888177724942891,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.791,
+      "step": 15469
+    },
+    {
+      "epoch": 0.7888687659503946,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 2.8191,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7889197594064999,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 15471
+    },
+    {
+      "epoch": 0.7889707528626053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 2.8418,
+      "step": 15472
+    },
+    {
+      "epoch": 0.7890217463187107,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8282,
+      "step": 15473
+    },
+    {
+      "epoch": 0.7890727397748161,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8297,
+      "step": 15474
+    },
+    {
+      "epoch": 0.7891237332309214,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8373,
+      "step": 15475
+    },
+    {
+      "epoch": 0.7891747266870268,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7968,
+      "step": 15476
+    },
+    {
+      "epoch": 0.7892257201431323,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8203,
+      "step": 15477
+    },
+    {
+      "epoch": 0.7892767135992377,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8522,
+      "step": 15478
+    },
+    {
+      "epoch": 0.789327707055343,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8193,
+      "step": 15479
+    },
+    {
+      "epoch": 0.7893787005114484,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8187,
+      "step": 15480
+    },
+    {
+      "epoch": 0.7894296939675538,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8215,
+      "step": 15481
+    },
+    {
+      "epoch": 0.7894806874236592,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8121,
+      "step": 15482
+    },
+    {
+      "epoch": 0.7895316808797646,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8244,
+      "step": 15483
+    },
+    {
+      "epoch": 0.78958267433587,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8451,
+      "step": 15484
+    },
+    {
+      "epoch": 0.7896336677919754,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8249,
+      "step": 15485
+    },
+    {
+      "epoch": 0.7896846612480808,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.7914,
+      "step": 15486
+    },
+    {
+      "epoch": 0.7897356547041862,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8194,
+      "step": 15487
+    },
+    {
+      "epoch": 0.7897866481602915,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8022,
+      "step": 15488
+    },
+    {
+      "epoch": 0.7898376416163969,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.7889,
+      "step": 15489
+    },
+    {
+      "epoch": 0.7898886350725023,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8276,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7899396285286078,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8353,
+      "step": 15491
+    },
+    {
+      "epoch": 0.7899906219847131,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 15492
+    },
+    {
+      "epoch": 0.7900416154408185,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.7943,
+      "step": 15493
+    },
+    {
+      "epoch": 0.7900926088969239,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 15494
+    },
+    {
+      "epoch": 0.7901436023530293,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8207,
+      "step": 15495
+    },
+    {
+      "epoch": 0.7901945958091346,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8281,
+      "step": 15496
+    },
+    {
+      "epoch": 0.79024558926524,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 15497
+    },
+    {
+      "epoch": 0.7902965827213455,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7877,
+      "step": 15498
+    },
+    {
+      "epoch": 0.7903475761774509,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.7956,
+      "step": 15499
+    },
+    {
+      "epoch": 0.7903985696335563,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8199,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7904495630896616,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.7959,
+      "step": 15501
+    },
+    {
+      "epoch": 0.790500556545767,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.845,
+      "step": 15502
+    },
+    {
+      "epoch": 0.7905515500018724,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 15503
+    },
+    {
+      "epoch": 0.7906025434579778,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7949,
+      "step": 15504
+    },
+    {
+      "epoch": 0.7906535369140832,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.002,
+      "loss": 2.8078,
+      "step": 15505
+    },
+    {
+      "epoch": 0.7907045303701886,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.7842,
+      "step": 15506
+    },
+    {
+      "epoch": 0.790755523826294,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8288,
+      "step": 15507
+    },
+    {
+      "epoch": 0.7908065172823994,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.814,
+      "step": 15508
+    },
+    {
+      "epoch": 0.7908575107385047,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.7966,
+      "step": 15509
+    },
+    {
+      "epoch": 0.7909085041946101,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.84,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7909594976507155,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 2.793,
+      "step": 15511
+    },
+    {
+      "epoch": 0.791010491106821,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 2.8103,
+      "step": 15512
+    },
+    {
+      "epoch": 0.7910614845629264,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 15513
+    },
+    {
+      "epoch": 0.7911124780190317,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 2.8214,
+      "step": 15514
+    },
+    {
+      "epoch": 0.7911634714751371,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.7632,
+      "step": 15515
+    },
+    {
+      "epoch": 0.7912144649312425,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.7748,
+      "step": 15516
+    },
+    {
+      "epoch": 0.7912654583873479,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 2.8168,
+      "step": 15517
+    },
+    {
+      "epoch": 0.7913164518434532,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.8416,
+      "step": 15518
+    },
+    {
+      "epoch": 0.7913674452995587,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.7963,
+      "step": 15519
+    },
+    {
+      "epoch": 0.7914184387556641,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8183,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7914694322117695,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8319,
+      "step": 15521
+    },
+    {
+      "epoch": 0.7915204256678748,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 15522
+    },
+    {
+      "epoch": 0.7915714191239802,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7979,
+      "step": 15523
+    },
+    {
+      "epoch": 0.7916224125800856,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 15524
+    },
+    {
+      "epoch": 0.791673406036191,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8074,
+      "step": 15525
+    },
+    {
+      "epoch": 0.7917243994922964,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 15526
+    },
+    {
+      "epoch": 0.7917753929484018,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 15527
+    },
+    {
+      "epoch": 0.7918263864045072,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8228,
+      "step": 15528
+    },
+    {
+      "epoch": 0.7918773798606126,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 2.8124,
+      "step": 15529
+    },
+    {
+      "epoch": 0.791928373316718,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8141,
+      "step": 15530
+    },
+    {
+      "epoch": 0.7919793667728233,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.7994,
+      "step": 15531
+    },
+    {
+      "epoch": 0.7920303602289287,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8252,
+      "step": 15532
+    },
+    {
+      "epoch": 0.7920813536850342,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 2.8262,
+      "step": 15533
+    },
+    {
+      "epoch": 0.7921323471411396,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.8212,
+      "step": 15534
+    },
+    {
+      "epoch": 0.7921833405972449,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8271,
+      "step": 15535
+    },
+    {
+      "epoch": 0.7922343340533503,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 15536
+    },
+    {
+      "epoch": 0.7922853275094557,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 15537
+    },
+    {
+      "epoch": 0.7923363209655611,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.8036,
+      "step": 15538
+    },
+    {
+      "epoch": 0.7923873144216664,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.8092,
+      "step": 15539
+    },
+    {
+      "epoch": 0.7924383078777719,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7924893013338773,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8195,
+      "step": 15541
+    },
+    {
+      "epoch": 0.7925402947899827,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 2.8197,
+      "step": 15542
+    },
+    {
+      "epoch": 0.792591288246088,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8436,
+      "step": 15543
+    },
+    {
+      "epoch": 0.7926422817021934,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.7875,
+      "step": 15544
+    },
+    {
+      "epoch": 0.7926932751582988,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 2.8064,
+      "step": 15545
+    },
+    {
+      "epoch": 0.7927442686144042,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8107,
+      "step": 15546
+    },
+    {
+      "epoch": 0.7927952620705097,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8016,
+      "step": 15547
+    },
+    {
+      "epoch": 0.792846255526615,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8073,
+      "step": 15548
+    },
+    {
+      "epoch": 0.7928972489827204,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 2.8018,
+      "step": 15549
+    },
+    {
+      "epoch": 0.7929482424388258,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 2.8196,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7929992358949312,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 2.8136,
+      "step": 15551
+    },
+    {
+      "epoch": 0.7930502293510365,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 2.8127,
+      "step": 15552
+    },
+    {
+      "epoch": 0.7931012228071419,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 15553
+    },
+    {
+      "epoch": 0.7931522162632474,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.797,
+      "step": 15554
+    },
+    {
+      "epoch": 0.7932032097193528,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 2.8231,
+      "step": 15555
+    },
+    {
+      "epoch": 0.7932542031754581,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 2.8118,
+      "step": 15556
+    },
+    {
+      "epoch": 0.7933051966315635,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 2.8101,
+      "step": 15557
+    },
+    {
+      "epoch": 0.7933561900876689,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8202,
+      "step": 15558
+    },
+    {
+      "epoch": 0.7934071835437743,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 2.8041,
+      "step": 15559
+    },
+    {
+      "epoch": 0.7934581769998796,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 2.8225,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7935091704559851,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8098,
+      "step": 15561
+    },
+    {
+      "epoch": 0.7935601639120905,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 2.8162,
+      "step": 15562
+    },
+    {
+      "epoch": 0.7936111573681959,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8386,
+      "step": 15563
+    },
+    {
+      "epoch": 0.7936621508243012,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 2.8072,
+      "step": 15564
+    },
+    {
+      "epoch": 0.7937131442804066,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8192,
+      "step": 15565
+    },
+    {
+      "epoch": 0.793764137736512,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 15566
+    },
+    {
+      "epoch": 0.7938151311926174,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8387,
+      "step": 15567
+    },
+    {
+      "epoch": 0.7938661246487229,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8143,
+      "step": 15568
+    },
+    {
+      "epoch": 0.7939171181048282,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8038,
+      "step": 15569
+    },
+    {
+      "epoch": 0.7939681115609336,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8265,
+      "step": 15570
+    },
+    {
+      "epoch": 0.794019105017039,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8106,
+      "step": 15571
+    },
+    {
+      "epoch": 0.7940700984731444,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 2.8182,
+      "step": 15572
+    },
+    {
+      "epoch": 0.7941210919292497,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.805,
+      "step": 15573
+    },
+    {
+      "epoch": 0.7941720853853551,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 2.8039,
+      "step": 15574
+    },
+    {
+      "epoch": 0.7942230788414606,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.8102,
+      "step": 15575
+    },
+    {
+      "epoch": 0.794274072297566,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 2.8444,
+      "step": 15576
+    },
+    {
+      "epoch": 0.7943250657536713,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 2.8421,
+      "step": 15577
+    },
+    {
+      "epoch": 0.7943760592097767,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8296,
+      "step": 15578
+    },
+    {
+      "epoch": 0.7944270526658821,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 2.8114,
+      "step": 15579
+    },
+    {
+      "epoch": 0.7944780461219875,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8307,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7945290395780928,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 2.8109,
+      "step": 15581
+    },
+    {
+      "epoch": 0.7945800330341983,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.8156,
+      "step": 15582
+    },
+    {
+      "epoch": 0.7946310264903037,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 2.8095,
+      "step": 15583
+    },
+    {
+      "epoch": 0.7946820199464091,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 2.7819,
+      "step": 15584
+    },
+    {
+      "epoch": 0.7947330134025145,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 2.8236,
+      "step": 15585
+    },
+    {
+      "epoch": 0.7947840068586198,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 2.7974,
+      "step": 15586
+    },
+    {
+      "epoch": 0.7948350003147252,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 2.8087,
+      "step": 15587
+    },
+    {
+      "epoch": 0.7948859937708306,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 2.8292,
+      "step": 15588
+    },
+    {
+      "epoch": 0.7949369872269361,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 2.7902,
+      "step": 15589
+    },
+    {
+      "epoch": 0.7949879806830414,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 2.8213,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7950389741391468,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8033,
+      "step": 15591
+    },
+    {
+      "epoch": 0.7950899675952522,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 15592
+    },
+    {
+      "epoch": 0.7951409610513576,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 2.801,
+      "step": 15593
+    },
+    {
+      "epoch": 0.7951919545074629,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 2.8113,
+      "step": 15594
+    },
+    {
+      "epoch": 0.7952429479635683,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 2.7977,
+      "step": 15595
+    },
+    {
+      "epoch": 0.7952939414196738,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 2.8412,
+      "step": 15596
+    },
+    {
+      "epoch": 0.7953449348757792,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.8263,
+      "step": 15597
+    },
+    {
+      "epoch": 0.7953959283318845,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8381,
+      "step": 15598
+    },
+    {
+      "epoch": 0.7954469217879899,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 2.836,
+      "step": 15599
+    },
+    {
+      "epoch": 0.7954979152440953,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.7948,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7955489087002007,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.8066,
+      "step": 15601
+    },
+    {
+      "epoch": 0.795599902156306,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 2.818,
+      "step": 15602
+    },
+    {
+      "epoch": 0.7956508956124115,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.81,
+      "step": 15603
+    },
+    {
+      "epoch": 0.7957018890685169,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 2.795,
+      "step": 15604
+    },
+    {
+      "epoch": 0.7957528825246223,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 2.8129,
+      "step": 15605
+    },
+    {
+      "epoch": 0.7958038759807277,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 2.7783,
+      "step": 15606
+    },
+    {
+      "epoch": 0.795854869436833,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 2.8391,
+      "step": 15607
+    },
+    {
+      "epoch": 0.7959058628929384,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 2.7932,
+      "step": 15608
+    },
+    {
+      "epoch": 0.7959568563490438,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 2.8111,
+      "step": 15609
+    },
+    {
+      "epoch": 0.7960078498051493,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 2.8245,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7960588432612546,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0019999997224173904,
+      "loss": 2.8289,
+      "step": 15611
+    },
+    {
+      "epoch": 0.79610983671736,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001999998889669733,
+      "loss": 2.8088,
+      "step": 15612
+    },
+    {
+      "epoch": 0.7961608301734654,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001999997501757542,
+      "loss": 2.8258,
+      "step": 15613
+    },
+    {
+      "epoch": 0.7962118236295708,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0019999955586816726,
+      "loss": 2.8165,
+      "step": 15614
+    },
+    {
+      "epoch": 0.7962628170856761,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0019999930604433238,
+      "loss": 2.8041,
+      "step": 15615
+    },
+    {
+      "epoch": 0.7963138105417815,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019999900070440365,
+      "loss": 2.8095,
+      "step": 15616
+    },
+    {
+      "epoch": 0.796364803997887,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.001999986398485694,
+      "loss": 2.8052,
+      "step": 15617
+    },
+    {
+      "epoch": 0.7964157974539924,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019999822347705233,
+      "loss": 2.8295,
+      "step": 15618
+    },
+    {
+      "epoch": 0.7964667909100978,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001999977515901092,
+      "loss": 2.813,
+      "step": 15619
+    },
+    {
+      "epoch": 0.7965177843662031,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001999972241880311,
+      "loss": 2.8169,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7965687778223085,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019999664127114333,
+      "loss": 2.8325,
+      "step": 15621
+    },
+    {
+      "epoch": 0.7966197712784139,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019999600283980546,
+      "loss": 2.8188,
+      "step": 15622
+    },
+    {
+      "epoch": 0.7966707647345193,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001999953088944114,
+      "loss": 2.8203,
+      "step": 15623
+    },
+    {
+      "epoch": 0.7967217581906247,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001999945594353891,
+      "loss": 2.8355,
+      "step": 15624
+    },
+    {
+      "epoch": 0.7967727516467301,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019999375446320094,
+      "loss": 2.7846,
+      "step": 15625
+    },
+    {
+      "epoch": 0.7968237451028355,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0019999289397834344,
+      "loss": 2.8084,
+      "step": 15626
+    },
+    {
+      "epoch": 0.7968747385589409,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001999919779813474,
+      "loss": 2.8094,
+      "step": 15627
+    },
+    {
+      "epoch": 0.7969257320150462,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001999910064727778,
+      "loss": 2.7995,
+      "step": 15628
+    },
+    {
+      "epoch": 0.7969767254711516,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00199989979453234,
+      "loss": 2.8253,
+      "step": 15629
+    },
+    {
+      "epoch": 0.797027718927257,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019998889692334947,
+      "loss": 2.8239,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7970787123833625,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0019998775888379196,
+      "loss": 2.8042,
+      "step": 15631
+    },
+    {
+      "epoch": 0.7971297058394679,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019998656533526346,
+      "loss": 2.7984,
+      "step": 15632
+    },
+    {
+      "epoch": 0.7971806992955732,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019998531627850027,
+      "loss": 2.8093,
+      "step": 15633
+    },
+    {
+      "epoch": 0.7972316927516786,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001999840117142728,
+      "loss": 2.814,
+      "step": 15634
+    },
+    {
+      "epoch": 0.797282686207784,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019998265164338585,
+      "loss": 2.8337,
+      "step": 15635
+    },
+    {
+      "epoch": 0.7973336796638893,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001999812360666783,
+      "loss": 2.8082,
+      "step": 15636
+    },
+    {
+      "epoch": 0.7973846731199947,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019997976498502337,
+      "loss": 2.8194,
+      "step": 15637
+    },
+    {
+      "epoch": 0.7974356665761002,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001999782383993286,
+      "loss": 2.8045,
+      "step": 15638
+    },
+    {
+      "epoch": 0.7974866600322056,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.001999766563105355,
+      "loss": 2.8437,
+      "step": 15639
+    },
+    {
+      "epoch": 0.797537653488311,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0019997501871962006,
+      "loss": 2.8271,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7975886469444163,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001999733256275924,
+      "loss": 2.7906,
+      "step": 15641
+    },
+    {
+      "epoch": 0.7976396404005217,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00199971577035497,
+      "loss": 2.8266,
+      "step": 15642
+    },
+    {
+      "epoch": 0.7976906338566271,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001999697729444124,
+      "loss": 2.799,
+      "step": 15643
+    },
+    {
+      "epoch": 0.7977416273127325,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001999679133554514,
+      "loss": 2.8008,
+      "step": 15644
+    },
+    {
+      "epoch": 0.797792620768838,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0019996599826976125,
+      "loss": 2.8078,
+      "step": 15645
+    },
+    {
+      "epoch": 0.7978436142249433,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001999640276885231,
+      "loss": 2.8054,
+      "step": 15646
+    },
+    {
+      "epoch": 0.7978946076810487,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001999620016129526,
+      "loss": 2.8325,
+      "step": 15647
+    },
+    {
+      "epoch": 0.7979456011371541,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001999599200442995,
+      "loss": 2.831,
+      "step": 15648
+    },
+    {
+      "epoch": 0.7979965945932594,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019995778298384785,
+      "loss": 2.805,
+      "step": 15649
+    },
+    {
+      "epoch": 0.7980475880493648,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019995559043291585,
+      "loss": 2.7701,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7980985815054702,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00199953342392856,
+      "loss": 2.822,
+      "step": 15651
+    },
+    {
+      "epoch": 0.7981495749615757,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00199951038865055,
+      "loss": 2.8186,
+      "step": 15652
+    },
+    {
+      "epoch": 0.7982005684176811,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001999486798509338,
+      "loss": 2.8033,
+      "step": 15653
+    },
+    {
+      "epoch": 0.7982515618737864,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019994626535194755,
+      "loss": 2.7926,
+      "step": 15654
+    },
+    {
+      "epoch": 0.7983025553298918,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019994379536958564,
+      "loss": 2.8236,
+      "step": 15655
+    },
+    {
+      "epoch": 0.7983535487859972,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019994126990537163,
+      "loss": 2.8079,
+      "step": 15656
+    },
+    {
+      "epoch": 0.7984045422421026,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019993868896086343,
+      "loss": 2.8026,
+      "step": 15657
+    },
+    {
+      "epoch": 0.7984555356982079,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019993605253765304,
+      "loss": 2.7934,
+      "step": 15658
+    },
+    {
+      "epoch": 0.7985065291543134,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019993336063736677,
+      "loss": 2.7961,
+      "step": 15659
+    },
+    {
+      "epoch": 0.7985575226104188,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001999306132616651,
+      "loss": 2.8404,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7986085160665242,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019992781041224274,
+      "loss": 2.7984,
+      "step": 15661
+    },
+    {
+      "epoch": 0.7986595095226295,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001999249520908286,
+      "loss": 2.7755,
+      "step": 15662
+    },
+    {
+      "epoch": 0.7987105029787349,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019992203829918595,
+      "loss": 2.8223,
+      "step": 15663
+    },
+    {
+      "epoch": 0.7987614964348403,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019991906903911207,
+      "loss": 2.7833,
+      "step": 15664
+    },
+    {
+      "epoch": 0.7988124898909457,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019991604431243855,
+      "loss": 2.8248,
+      "step": 15665
+    },
+    {
+      "epoch": 0.7988634833470512,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019991296412103124,
+      "loss": 2.775,
+      "step": 15666
+    },
+    {
+      "epoch": 0.7989144768031565,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019990982846679007,
+      "loss": 2.7934,
+      "step": 15667
+    },
+    {
+      "epoch": 0.7989654702592619,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001999066373516494,
+      "loss": 2.8315,
+      "step": 15668
+    },
+    {
+      "epoch": 0.7990164637153673,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001999033907775776,
+      "loss": 2.8296,
+      "step": 15669
+    },
+    {
+      "epoch": 0.7990674571714727,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001999000887465773,
+      "loss": 2.7999,
+      "step": 15670
+    },
+    {
+      "epoch": 0.799118450627578,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019989673126068535,
+      "loss": 2.8107,
+      "step": 15671
+    },
+    {
+      "epoch": 0.7991694440836834,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001998933183219729,
+      "loss": 2.8141,
+      "step": 15672
+    },
+    {
+      "epoch": 0.7992204375397889,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019988984993254515,
+      "loss": 2.8226,
+      "step": 15673
+    },
+    {
+      "epoch": 0.7992714309958943,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001998863260945416,
+      "loss": 2.7926,
+      "step": 15674
+    },
+    {
+      "epoch": 0.7993224244519996,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001998827468101359,
+      "loss": 2.8334,
+      "step": 15675
+    },
+    {
+      "epoch": 0.799373417908105,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019987911208153596,
+      "loss": 2.8076,
+      "step": 15676
+    },
+    {
+      "epoch": 0.7994244113642104,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019987542191098385,
+      "loss": 2.8205,
+      "step": 15677
+    },
+    {
+      "epoch": 0.7994754048203158,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.001998716763007559,
+      "loss": 2.8472,
+      "step": 15678
+    },
+    {
+      "epoch": 0.7995263982764211,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019986787525316254,
+      "loss": 2.8202,
+      "step": 15679
+    },
+    {
+      "epoch": 0.7995773917325266,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019986401877054843,
+      "loss": 2.8074,
+      "step": 15680
+    },
+    {
+      "epoch": 0.799628385188632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0019986010685529246,
+      "loss": 2.7972,
+      "step": 15681
+    },
+    {
+      "epoch": 0.7996793786447374,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019985613950980778,
+      "loss": 2.8314,
+      "step": 15682
+    },
+    {
+      "epoch": 0.7997303721008427,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.001998521167365415,
+      "loss": 2.8558,
+      "step": 15683
+    },
+    {
+      "epoch": 0.7997813655569481,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001998480385379752,
+      "loss": 2.7985,
+      "step": 15684
+    },
+    {
+      "epoch": 0.7998323590130535,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019984390491662447,
+      "loss": 2.792,
+      "step": 15685
+    },
+    {
+      "epoch": 0.7998833524691589,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019983971587503907,
+      "loss": 2.814,
+      "step": 15686
+    },
+    {
+      "epoch": 0.7999343459252644,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001998354714158031,
+      "loss": 2.799,
+      "step": 15687
+    },
+    {
+      "epoch": 0.7999853393813697,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.001998311715415347,
+      "loss": 2.821,
+      "step": 15688
+    },
+    {
+      "epoch": 0.8000363328374751,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001998268162548863,
+      "loss": 2.7931,
+      "step": 15689
+    },
+    {
+      "epoch": 0.8000873262935805,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0019982240555854445,
+      "loss": 2.8048,
+      "step": 15690
+    },
+    {
+      "epoch": 0.8001383197496859,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019981793945522985,
+      "loss": 2.8408,
+      "step": 15691
+    },
+    {
+      "epoch": 0.8001893132057912,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019981341794769744,
+      "loss": 2.825,
+      "step": 15692
+    },
+    {
+      "epoch": 0.8002403066618966,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019980884103873633,
+      "loss": 2.8097,
+      "step": 15693
+    },
+    {
+      "epoch": 0.8002913001180021,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0019980420873116976,
+      "loss": 2.8132,
+      "step": 15694
+    },
+    {
+      "epoch": 0.8003422935741075,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0019979952102785513,
+      "loss": 2.82,
+      "step": 15695
+    },
+    {
+      "epoch": 0.8003932870302128,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001997947779316842,
+      "loss": 2.8082,
+      "step": 15696
+    },
+    {
+      "epoch": 0.8004442804863182,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019978997944558254,
+      "loss": 2.7776,
+      "step": 15697
+    },
+    {
+      "epoch": 0.8004952739424236,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019978512557251027,
+      "loss": 2.8315,
+      "step": 15698
+    },
+    {
+      "epoch": 0.800546267398529,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019978021631546143,
+      "loss": 2.8138,
+      "step": 15699
+    },
+    {
+      "epoch": 0.8005972608546343,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0019977525167746433,
+      "loss": 2.8248,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8006482543107398,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001997702316615813,
+      "loss": 2.8161,
+      "step": 15701
+    },
+    {
+      "epoch": 0.8006992477668452,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019976515627090916,
+      "loss": 2.8057,
+      "step": 15702
+    },
+    {
+      "epoch": 0.8007502412229506,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019976002550857845,
+      "loss": 2.8201,
+      "step": 15703
+    },
+    {
+      "epoch": 0.800801234679056,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019975483937775414,
+      "loss": 2.8261,
+      "step": 15704
+    },
+    {
+      "epoch": 0.8008522281351613,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019974959788163537,
+      "loss": 2.8251,
+      "step": 15705
+    },
+    {
+      "epoch": 0.8009032215912667,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019974430102345526,
+      "loss": 2.8149,
+      "step": 15706
+    },
+    {
+      "epoch": 0.8009542150473721,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.001997389488064812,
+      "loss": 2.8042,
+      "step": 15707
+    },
+    {
+      "epoch": 0.8010052085034776,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001997335412340148,
+      "loss": 2.8072,
+      "step": 15708
+    },
+    {
+      "epoch": 0.8010562019595829,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019972807830939163,
+      "loss": 2.797,
+      "step": 15709
+    },
+    {
+      "epoch": 0.8011071954156883,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019972256003598153,
+      "loss": 2.7923,
+      "step": 15710
+    },
+    {
+      "epoch": 0.8011581888717937,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019971698641718843,
+      "loss": 2.8236,
+      "step": 15711
+    },
+    {
+      "epoch": 0.8012091823278991,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001997113574564504,
+      "loss": 2.7981,
+      "step": 15712
+    },
+    {
+      "epoch": 0.8012601757840044,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001997056731572397,
+      "loss": 2.7775,
+      "step": 15713
+    },
+    {
+      "epoch": 0.8013111692401098,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001996999335230627,
+      "loss": 2.8266,
+      "step": 15714
+    },
+    {
+      "epoch": 0.8013621626962153,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019969413855745986,
+      "loss": 2.8435,
+      "step": 15715
+    },
+    {
+      "epoch": 0.8014131561523207,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019968828826400584,
+      "loss": 2.8071,
+      "step": 15716
+    },
+    {
+      "epoch": 0.801464149608426,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019968238264630937,
+      "loss": 2.802,
+      "step": 15717
+    },
+    {
+      "epoch": 0.8015151430645314,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.001996764217080133,
+      "loss": 2.8202,
+      "step": 15718
+    },
+    {
+      "epoch": 0.8015661365206368,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0019967040545279474,
+      "loss": 2.8068,
+      "step": 15719
+    },
+    {
+      "epoch": 0.8016171299767422,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001996643338843647,
+      "loss": 2.8127,
+      "step": 15720
+    },
+    {
+      "epoch": 0.8016681234328475,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001996582070064685,
+      "loss": 2.8464,
+      "step": 15721
+    },
+    {
+      "epoch": 0.801719116888953,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019965202482288553,
+      "loss": 2.8322,
+      "step": 15722
+    },
+    {
+      "epoch": 0.8017701103450584,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001996457873374292,
+      "loss": 2.8296,
+      "step": 15723
+    },
+    {
+      "epoch": 0.8018211038011638,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0019963949455394712,
+      "loss": 2.8073,
+      "step": 15724
+    },
+    {
+      "epoch": 0.8018720972572692,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00199633146476321,
+      "loss": 2.8246,
+      "step": 15725
+    },
+    {
+      "epoch": 0.8019230907133745,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.001996267431084667,
+      "loss": 2.8184,
+      "step": 15726
+    },
+    {
+      "epoch": 0.8019740841694799,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001996202844543341,
+      "loss": 2.8204,
+      "step": 15727
+    },
+    {
+      "epoch": 0.8020250776255853,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0019961377051790725,
+      "loss": 2.8116,
+      "step": 15728
+    },
+    {
+      "epoch": 0.8020760710816908,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001996072013032042,
+      "loss": 2.8231,
+      "step": 15729
+    },
+    {
+      "epoch": 0.8021270645377961,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001996005768142772,
+      "loss": 2.8129,
+      "step": 15730
+    },
+    {
+      "epoch": 0.8021780579939015,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001995938970552126,
+      "loss": 2.7948,
+      "step": 15731
+    },
+    {
+      "epoch": 0.8022290514500069,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0019958716203013083,
+      "loss": 2.8249,
+      "step": 15732
+    },
+    {
+      "epoch": 0.8022800449061123,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001995803717431864,
+      "loss": 2.8136,
+      "step": 15733
+    },
+    {
+      "epoch": 0.8023310383622176,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019957352619856777,
+      "loss": 2.7833,
+      "step": 15734
+    },
+    {
+      "epoch": 0.802382031818323,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0019956662540049773,
+      "loss": 2.8122,
+      "step": 15735
+    },
+    {
+      "epoch": 0.8024330252744285,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0019955966935323306,
+      "loss": 2.8191,
+      "step": 15736
+    },
+    {
+      "epoch": 0.8024840187305339,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019955265806106447,
+      "loss": 2.8221,
+      "step": 15737
+    },
+    {
+      "epoch": 0.8025350121866393,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0019954559152831706,
+      "loss": 2.7979,
+      "step": 15738
+    },
+    {
+      "epoch": 0.8025860056427446,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001995384697593496,
+      "loss": 2.7902,
+      "step": 15739
+    },
+    {
+      "epoch": 0.80263699909885,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001995312927585554,
+      "loss": 2.8143,
+      "step": 15740
+    },
+    {
+      "epoch": 0.8026879925549554,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001995240605303614,
+      "loss": 2.8232,
+      "step": 15741
+    },
+    {
+      "epoch": 0.8027389860110608,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001995167730792289,
+      "loss": 2.8514,
+      "step": 15742
+    },
+    {
+      "epoch": 0.8027899794671662,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019950943040965316,
+      "loss": 2.8009,
+      "step": 15743
+    },
+    {
+      "epoch": 0.8028409729232716,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019950203252616346,
+      "loss": 2.8371,
+      "step": 15744
+    },
+    {
+      "epoch": 0.802891966379377,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019949457943332326,
+      "loss": 2.8205,
+      "step": 15745
+    },
+    {
+      "epoch": 0.8029429598354824,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001994870711357299,
+      "loss": 2.812,
+      "step": 15746
+    },
+    {
+      "epoch": 0.8029939532915877,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019947950763801498,
+      "loss": 2.806,
+      "step": 15747
+    },
+    {
+      "epoch": 0.8030449467476931,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019947188894484397,
+      "loss": 2.8025,
+      "step": 15748
+    },
+    {
+      "epoch": 0.8030959402037985,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001994642150609165,
+      "loss": 2.8314,
+      "step": 15749
+    },
+    {
+      "epoch": 0.803146933659904,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001994564859909662,
+      "loss": 2.796,
+      "step": 15750
+    },
+    {
+      "epoch": 0.8031979271160093,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019944870173976073,
+      "loss": 2.8326,
+      "step": 15751
+    },
+    {
+      "epoch": 0.8032489205721147,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001994408623121018,
+      "loss": 2.8072,
+      "step": 15752
+    },
+    {
+      "epoch": 0.8032999140282201,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019943296771282523,
+      "loss": 2.817,
+      "step": 15753
+    },
+    {
+      "epoch": 0.8033509074843255,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019942501794680077,
+      "loss": 2.8071,
+      "step": 15754
+    },
+    {
+      "epoch": 0.8034019009404308,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001994170130189322,
+      "loss": 2.8233,
+      "step": 15755
+    },
+    {
+      "epoch": 0.8034528943965362,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001994089529341574,
+      "loss": 2.825,
+      "step": 15756
+    },
+    {
+      "epoch": 0.8035038878526417,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019940083769744823,
+      "loss": 2.7968,
+      "step": 15757
+    },
+    {
+      "epoch": 0.8035548813087471,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001993926673138105,
+      "loss": 2.8241,
+      "step": 15758
+    },
+    {
+      "epoch": 0.8036058747648525,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001993844417882843,
+      "loss": 2.835,
+      "step": 15759
+    },
+    {
+      "epoch": 0.8036568682209578,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001993761611259434,
+      "loss": 2.7859,
+      "step": 15760
+    },
+    {
+      "epoch": 0.8037078616770632,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001993678253318957,
+      "loss": 2.8161,
+      "step": 15761
+    },
+    {
+      "epoch": 0.8037588551331686,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0019935943441128324,
+      "loss": 2.8269,
+      "step": 15762
+    },
+    {
+      "epoch": 0.803809848589274,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.001993509883692819,
+      "loss": 2.8102,
+      "step": 15763
+    },
+    {
+      "epoch": 0.8038608420453794,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0019934248721110164,
+      "loss": 2.8221,
+      "step": 15764
+    },
+    {
+      "epoch": 0.8039118355014848,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.001993339309419864,
+      "loss": 2.8526,
+      "step": 15765
+    },
+    {
+      "epoch": 0.8039628289575902,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019932531956721416,
+      "loss": 2.81,
+      "step": 15766
+    },
+    {
+      "epoch": 0.8040138224136956,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0019931665309209676,
+      "loss": 2.8129,
+      "step": 15767
+    },
+    {
+      "epoch": 0.8040648158698009,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019930793152198017,
+      "loss": 2.7903,
+      "step": 15768
+    },
+    {
+      "epoch": 0.8041158093259063,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.001992991548622443,
+      "loss": 2.8115,
+      "step": 15769
+    },
+    {
+      "epoch": 0.8041668027820117,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00199290323118303,
+      "loss": 2.818,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8042177962381172,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001992814362956042,
+      "loss": 2.7745,
+      "step": 15771
+    },
+    {
+      "epoch": 0.8042687896942226,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019927249439962965,
+      "loss": 2.8348,
+      "step": 15772
+    },
+    {
+      "epoch": 0.8043197831503279,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001992634974358952,
+      "loss": 2.7943,
+      "step": 15773
+    },
+    {
+      "epoch": 0.8043707766064333,
+      "grad_norm": 0.625,
+      "learning_rate": 0.001992544454099507,
+      "loss": 2.8093,
+      "step": 15774
+    },
+    {
+      "epoch": 0.8044217700625387,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001992453383273797,
+      "loss": 2.8244,
+      "step": 15775
+    },
+    {
+      "epoch": 0.804472763518644,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019923617619380016,
+      "loss": 2.8171,
+      "step": 15776
+    },
+    {
+      "epoch": 0.8045237569747494,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0019922695901486358,
+      "loss": 2.8333,
+      "step": 15777
+    },
+    {
+      "epoch": 0.8045747504308549,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001992176867962556,
+      "loss": 2.7966,
+      "step": 15778
+    },
+    {
+      "epoch": 0.8046257438869603,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001992083595436958,
+      "loss": 2.8299,
+      "step": 15779
+    },
+    {
+      "epoch": 0.8046767373430657,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0019919897726293776,
+      "loss": 2.8066,
+      "step": 15780
+    },
+    {
+      "epoch": 0.804727730799171,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019918953995976883,
+      "loss": 2.8103,
+      "step": 15781
+    },
+    {
+      "epoch": 0.8047787242552764,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001991800476400105,
+      "loss": 2.8096,
+      "step": 15782
+    },
+    {
+      "epoch": 0.8048297177113818,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019917050030951807,
+      "loss": 2.8259,
+      "step": 15783
+    },
+    {
+      "epoch": 0.8048807111674872,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001991608979741808,
+      "loss": 2.8015,
+      "step": 15784
+    },
+    {
+      "epoch": 0.8049317046235926,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0019915124063992194,
+      "loss": 2.8024,
+      "step": 15785
+    },
+    {
+      "epoch": 0.804982698079698,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001991415283126986,
+      "loss": 2.8106,
+      "step": 15786
+    },
+    {
+      "epoch": 0.8050336915358034,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019913176099850185,
+      "loss": 2.8194,
+      "step": 15787
+    },
+    {
+      "epoch": 0.8050846849919088,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.001991219387033566,
+      "loss": 2.819,
+      "step": 15788
+    },
+    {
+      "epoch": 0.8051356784480141,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.001991120614333218,
+      "loss": 2.8088,
+      "step": 15789
+    },
+    {
+      "epoch": 0.8051866719041195,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.001991021291944902,
+      "loss": 2.807,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8052376653602249,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019909214199298852,
+      "loss": 2.8606,
+      "step": 15791
+    },
+    {
+      "epoch": 0.8052886588163304,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019908209983497734,
+      "loss": 2.791,
+      "step": 15792
+    },
+    {
+      "epoch": 0.8053396522724358,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0019907200272665125,
+      "loss": 2.8051,
+      "step": 15793
+    },
+    {
+      "epoch": 0.8053906457285411,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.001990618506742386,
+      "loss": 2.8185,
+      "step": 15794
+    },
+    {
+      "epoch": 0.8054416391846465,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0019905164368400168,
+      "loss": 2.8081,
+      "step": 15795
+    },
+    {
+      "epoch": 0.8054926326407519,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019904138176223664,
+      "loss": 2.7918,
+      "step": 15796
+    },
+    {
+      "epoch": 0.8055436260968573,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0019903106491527367,
+      "loss": 2.7916,
+      "step": 15797
+    },
+    {
+      "epoch": 0.8055946195529626,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019902069314947654,
+      "loss": 2.7818,
+      "step": 15798
+    },
+    {
+      "epoch": 0.8056456130090681,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019901026647124328,
+      "loss": 2.8207,
+      "step": 15799
+    },
+    {
+      "epoch": 0.8056966064651735,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019899978488700544,
+      "loss": 2.8599,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8057475999212789,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001989892484032286,
+      "loss": 2.7851,
+      "step": 15801
+    },
+    {
+      "epoch": 0.8057985933773842,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019897865702641227,
+      "loss": 2.8113,
+      "step": 15802
+    },
+    {
+      "epoch": 0.8058495868334896,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.001989680107630897,
+      "loss": 2.8051,
+      "step": 15803
+    },
+    {
+      "epoch": 0.805900580289595,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019895730961982804,
+      "loss": 2.8241,
+      "step": 15804
+    },
+    {
+      "epoch": 0.8059515737457004,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0019894655360322825,
+      "loss": 2.8246,
+      "step": 15805
+    },
+    {
+      "epoch": 0.8060025672018059,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019893574271992527,
+      "loss": 2.816,
+      "step": 15806
+    },
+    {
+      "epoch": 0.8060535606579112,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.001989248769765877,
+      "loss": 2.7871,
+      "step": 15807
+    },
+    {
+      "epoch": 0.8061045541140166,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001989139563799182,
+      "loss": 2.8283,
+      "step": 15808
+    },
+    {
+      "epoch": 0.806155547570122,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00198902980936653,
+      "loss": 2.8106,
+      "step": 15809
+    },
+    {
+      "epoch": 0.8062065410262274,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019889195065356242,
+      "loss": 2.7957,
+      "step": 15810
+    },
+    {
+      "epoch": 0.8062575344823327,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001988808655374504,
+      "loss": 2.8121,
+      "step": 15811
+    },
+    {
+      "epoch": 0.8063085279384381,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001988697255951549,
+      "loss": 2.8133,
+      "step": 15812
+    },
+    {
+      "epoch": 0.8063595213945436,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001988585308335475,
+      "loss": 2.7999,
+      "step": 15813
+    },
+    {
+      "epoch": 0.806410514850649,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001988472812595337,
+      "loss": 2.7949,
+      "step": 15814
+    },
+    {
+      "epoch": 0.8064615083067543,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001988359768800529,
+      "loss": 2.8127,
+      "step": 15815
+    },
+    {
+      "epoch": 0.8065125017628597,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0019882461770207812,
+      "loss": 2.8081,
+      "step": 15816
+    },
+    {
+      "epoch": 0.8065634952189651,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001988132037326163,
+      "loss": 2.8106,
+      "step": 15817
+    },
+    {
+      "epoch": 0.8066144886750705,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.001988017349787081,
+      "loss": 2.7921,
+      "step": 15818
+    },
+    {
+      "epoch": 0.8066654821311758,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019879021144742813,
+      "loss": 2.8223,
+      "step": 15819
+    },
+    {
+      "epoch": 0.8067164755872813,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001987786331458846,
+      "loss": 2.8,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8067674690433867,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019876700008121958,
+      "loss": 2.8068,
+      "step": 15821
+    },
+    {
+      "epoch": 0.8068184624994921,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00198755312260609,
+      "loss": 2.8166,
+      "step": 15822
+    },
+    {
+      "epoch": 0.8068694559555974,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0019874356969126245,
+      "loss": 2.803,
+      "step": 15823
+    },
+    {
+      "epoch": 0.8069204494117028,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001987317723804233,
+      "loss": 2.8293,
+      "step": 15824
+    },
+    {
+      "epoch": 0.8069714428678082,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001987199203353688,
+      "loss": 2.8205,
+      "step": 15825
+    },
+    {
+      "epoch": 0.8070224363239136,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.001987080135634098,
+      "loss": 2.7886,
+      "step": 15826
+    },
+    {
+      "epoch": 0.8070734297800191,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001986960520718911,
+      "loss": 2.8183,
+      "step": 15827
+    },
+    {
+      "epoch": 0.8071244232361244,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0019868403586819107,
+      "loss": 2.8052,
+      "step": 15828
+    },
+    {
+      "epoch": 0.8071754166922298,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019867196495972194,
+      "loss": 2.812,
+      "step": 15829
+    },
+    {
+      "epoch": 0.8072264101483352,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019865983935392966,
+      "loss": 2.7987,
+      "step": 15830
+    },
+    {
+      "epoch": 0.8072774036044406,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001986476590582939,
+      "loss": 2.7832,
+      "step": 15831
+    },
+    {
+      "epoch": 0.8073283970605459,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00198635424080328,
+      "loss": 2.8328,
+      "step": 15832
+    },
+    {
+      "epoch": 0.8073793905166513,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001986231344275793,
+      "loss": 2.7974,
+      "step": 15833
+    },
+    {
+      "epoch": 0.8074303839727568,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.001986107901076285,
+      "loss": 2.7863,
+      "step": 15834
+    },
+    {
+      "epoch": 0.8074813774288622,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0019859839112809025,
+      "loss": 2.8155,
+      "step": 15835
+    },
+    {
+      "epoch": 0.8075323708849675,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001985859374966129,
+      "loss": 2.8011,
+      "step": 15836
+    },
+    {
+      "epoch": 0.8075833643410729,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001985734292208784,
+      "loss": 2.81,
+      "step": 15837
+    },
+    {
+      "epoch": 0.8076343577971783,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019856086630860256,
+      "loss": 2.8193,
+      "step": 15838
+    },
+    {
+      "epoch": 0.8076853512532837,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019854824876753482,
+      "loss": 2.8103,
+      "step": 15839
+    },
+    {
+      "epoch": 0.807736344709389,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019853557660545825,
+      "loss": 2.8154,
+      "step": 15840
+    },
+    {
+      "epoch": 0.8077873381654945,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001985228498301897,
+      "loss": 2.8261,
+      "step": 15841
+    },
+    {
+      "epoch": 0.8078383316215999,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.001985100684495797,
+      "loss": 2.795,
+      "step": 15842
+    },
+    {
+      "epoch": 0.8078893250777053,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001984972324715124,
+      "loss": 2.8275,
+      "step": 15843
+    },
+    {
+      "epoch": 0.8079403185338107,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0019848434190390574,
+      "loss": 2.8226,
+      "step": 15844
+    },
+    {
+      "epoch": 0.807991311989916,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019847139675471122,
+      "loss": 2.8196,
+      "step": 15845
+    },
+    {
+      "epoch": 0.8080423054460214,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0019845839703191413,
+      "loss": 2.8023,
+      "step": 15846
+    },
+    {
+      "epoch": 0.8080932989021268,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001984453427435333,
+      "loss": 2.812,
+      "step": 15847
+    },
+    {
+      "epoch": 0.8081442923582323,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019843223389762123,
+      "loss": 2.7963,
+      "step": 15848
+    },
+    {
+      "epoch": 0.8081952858143376,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019841907050226417,
+      "loss": 2.8117,
+      "step": 15849
+    },
+    {
+      "epoch": 0.808246279270443,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00198405852565582,
+      "loss": 2.8019,
+      "step": 15850
+    },
+    {
+      "epoch": 0.8082972727265484,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001983925800957281,
+      "loss": 2.8157,
+      "step": 15851
+    },
+    {
+      "epoch": 0.8083482661826538,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0019837925310088973,
+      "loss": 2.8251,
+      "step": 15852
+    },
+    {
+      "epoch": 0.8083992596387591,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.001983658715892875,
+      "loss": 2.814,
+      "step": 15853
+    },
+    {
+      "epoch": 0.8084502530948645,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001983524355691759,
+      "loss": 2.8074,
+      "step": 15854
+    },
+    {
+      "epoch": 0.80850124655097,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019833894504884294,
+      "loss": 2.8319,
+      "step": 15855
+    },
+    {
+      "epoch": 0.8085522400070754,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.001983254000366102,
+      "loss": 2.8166,
+      "step": 15856
+    },
+    {
+      "epoch": 0.8086032334631807,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0019831180054083295,
+      "loss": 2.8099,
+      "step": 15857
+    },
+    {
+      "epoch": 0.8086542269192861,
+      "grad_norm": 4.53125,
+      "learning_rate": 0.0019829814656990006,
+      "loss": 2.8609,
+      "step": 15858
+    },
+    {
+      "epoch": 0.8087052203753915,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0019828443813223397,
+      "loss": 2.8392,
+      "step": 15859
+    },
+    {
+      "epoch": 0.8087562138314969,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0019827067523629075,
+      "loss": 2.8744,
+      "step": 15860
+    },
+    {
+      "epoch": 0.8088072072876022,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0019825685789056,
+      "loss": 2.8161,
+      "step": 15861
+    },
+    {
+      "epoch": 0.8088582007437077,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00198242986103565,
+      "loss": 2.8689,
+      "step": 15862
+    },
+    {
+      "epoch": 0.8089091941998131,
+      "grad_norm": 25.875,
+      "learning_rate": 0.0019822905988386257,
+      "loss": 3.067,
+      "step": 15863
+    },
+    {
+      "epoch": 0.8089601876559185,
+      "grad_norm": 5.0625,
+      "learning_rate": 0.0019821507924004302,
+      "loss": 2.846,
+      "step": 15864
+    },
+    {
+      "epoch": 0.8090111811120239,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.001982010441807304,
+      "loss": 2.8272,
+      "step": 15865
+    },
+    {
+      "epoch": 0.8090621745681292,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.001981869547145822,
+      "loss": 2.8309,
+      "step": 15866
+    },
+    {
+      "epoch": 0.8091131680242346,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0019817281085028958,
+      "loss": 2.8507,
+      "step": 15867
+    },
+    {
+      "epoch": 0.80916416148034,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.001981586125965771,
+      "loss": 2.8027,
+      "step": 15868
+    },
+    {
+      "epoch": 0.8092151549364455,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0019814435996220294,
+      "loss": 2.8169,
+      "step": 15869
+    },
+    {
+      "epoch": 0.8092661483925508,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.001981300529559589,
+      "loss": 2.8484,
+      "step": 15870
+    },
+    {
+      "epoch": 0.8093171418486562,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0019811569158667028,
+      "loss": 2.8227,
+      "step": 15871
+    },
+    {
+      "epoch": 0.8093681353047616,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0019810127586319588,
+      "loss": 2.8477,
+      "step": 15872
+    },
+    {
+      "epoch": 0.809419128760867,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0019808680579442794,
+      "loss": 2.8075,
+      "step": 15873
+    },
+    {
+      "epoch": 0.8094701222169723,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001980722813892925,
+      "loss": 2.8417,
+      "step": 15874
+    },
+    {
+      "epoch": 0.8095211156730777,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.001980577026567488,
+      "loss": 2.7983,
+      "step": 15875
+    },
+    {
+      "epoch": 0.8095721091291832,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001980430696057898,
+      "loss": 2.8501,
+      "step": 15876
+    },
+    {
+      "epoch": 0.8096231025852886,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001980283822454419,
+      "loss": 2.8172,
+      "step": 15877
+    },
+    {
+      "epoch": 0.809674096041394,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0019801364058476497,
+      "loss": 2.821,
+      "step": 15878
+    },
+    {
+      "epoch": 0.8097250894974993,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019799884463285243,
+      "loss": 2.8621,
+      "step": 15879
+    },
+    {
+      "epoch": 0.8097760829536047,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0019798399439883117,
+      "loss": 2.8279,
+      "step": 15880
+    },
+    {
+      "epoch": 0.8098270764097101,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019796908989186154,
+      "loss": 2.822,
+      "step": 15881
+    },
+    {
+      "epoch": 0.8098780698658155,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019795413112113744,
+      "loss": 2.809,
+      "step": 15882
+    },
+    {
+      "epoch": 0.8099290633219209,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.001979391180958861,
+      "loss": 2.8208,
+      "step": 15883
+    },
+    {
+      "epoch": 0.8099800567780263,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.001979240508253684,
+      "loss": 2.7863,
+      "step": 15884
+    },
+    {
+      "epoch": 0.8100310502341317,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001979089293188785,
+      "loss": 2.8233,
+      "step": 15885
+    },
+    {
+      "epoch": 0.8100820436902371,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0019789375358574422,
+      "loss": 2.8323,
+      "step": 15886
+    },
+    {
+      "epoch": 0.8101330371463424,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0019787852363532662,
+      "loss": 2.8221,
+      "step": 15887
+    },
+    {
+      "epoch": 0.8101840306024478,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001978632394770203,
+      "loss": 2.8084,
+      "step": 15888
+    },
+    {
+      "epoch": 0.8102350240585532,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001978479011202534,
+      "loss": 2.8265,
+      "step": 15889
+    },
+    {
+      "epoch": 0.8102860175146587,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019783250857448725,
+      "loss": 2.7956,
+      "step": 15890
+    },
+    {
+      "epoch": 0.810337010970764,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019781706184921685,
+      "loss": 2.8066,
+      "step": 15891
+    },
+    {
+      "epoch": 0.8103880044268694,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001978015609539705,
+      "loss": 2.8002,
+      "step": 15892
+    },
+    {
+      "epoch": 0.8104389978829748,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019778600589830985,
+      "loss": 2.8123,
+      "step": 15893
+    },
+    {
+      "epoch": 0.8104899913390802,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001977703966918301,
+      "loss": 2.8265,
+      "step": 15894
+    },
+    {
+      "epoch": 0.8105409847951855,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001977547333441598,
+      "loss": 2.8467,
+      "step": 15895
+    },
+    {
+      "epoch": 0.8105919782512909,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0019773901586496094,
+      "loss": 2.8477,
+      "step": 15896
+    },
+    {
+      "epoch": 0.8106429717073964,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001977232442639287,
+      "loss": 2.8345,
+      "step": 15897
+    },
+    {
+      "epoch": 0.8106939651635018,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0019770741855079195,
+      "loss": 2.8026,
+      "step": 15898
+    },
+    {
+      "epoch": 0.8107449586196072,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001976915387353127,
+      "loss": 2.8583,
+      "step": 15899
+    },
+    {
+      "epoch": 0.8107959520757125,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019767560482728645,
+      "loss": 2.841,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8108469455318179,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019765961683654203,
+      "loss": 2.7891,
+      "step": 15901
+    },
+    {
+      "epoch": 0.8108979389879233,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.001976435747729416,
+      "loss": 2.7816,
+      "step": 15902
+    },
+    {
+      "epoch": 0.8109489324440287,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.001976274786463808,
+      "loss": 2.8002,
+      "step": 15903
+    },
+    {
+      "epoch": 0.8109999259001341,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.001976113284667884,
+      "loss": 2.8355,
+      "step": 15904
+    },
+    {
+      "epoch": 0.8110509193562395,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019759512424412672,
+      "loss": 2.8094,
+      "step": 15905
+    },
+    {
+      "epoch": 0.8111019128123449,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001975788659883914,
+      "loss": 2.8362,
+      "step": 15906
+    },
+    {
+      "epoch": 0.8111529062684503,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0019756255370961123,
+      "loss": 2.809,
+      "step": 15907
+    },
+    {
+      "epoch": 0.8112038997245556,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001975461874178485,
+      "loss": 2.8099,
+      "step": 15908
+    },
+    {
+      "epoch": 0.811254893180661,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001975297671231988,
+      "loss": 2.8204,
+      "step": 15909
+    },
+    {
+      "epoch": 0.8113058866367664,
+      "grad_norm": 7.46875,
+      "learning_rate": 0.001975132928357909,
+      "loss": 2.7933,
+      "step": 15910
+    },
+    {
+      "epoch": 0.8113568800928719,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0019749676456578703,
+      "loss": 2.841,
+      "step": 15911
+    },
+    {
+      "epoch": 0.8114078735489773,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001974801823233826,
+      "loss": 2.8112,
+      "step": 15912
+    },
+    {
+      "epoch": 0.8114588670050826,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001974635461188065,
+      "loss": 2.8442,
+      "step": 15913
+    },
+    {
+      "epoch": 0.811509860461188,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0019744685596232068,
+      "loss": 2.8105,
+      "step": 15914
+    },
+    {
+      "epoch": 0.8115608539172934,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019743011186422045,
+      "loss": 2.8092,
+      "step": 15915
+    },
+    {
+      "epoch": 0.8116118473733988,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019741331383483444,
+      "loss": 2.8249,
+      "step": 15916
+    },
+    {
+      "epoch": 0.8116628408295041,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001973964618845245,
+      "loss": 2.7926,
+      "step": 15917
+    },
+    {
+      "epoch": 0.8117138342856096,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001973795560236858,
+      "loss": 2.786,
+      "step": 15918
+    },
+    {
+      "epoch": 0.811764827741715,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001973625962627467,
+      "loss": 2.7978,
+      "step": 15919
+    },
+    {
+      "epoch": 0.8118158211978204,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0019734558261216882,
+      "loss": 2.8301,
+      "step": 15920
+    },
+    {
+      "epoch": 0.8118668146539257,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019732851508244705,
+      "loss": 2.8349,
+      "step": 15921
+    },
+    {
+      "epoch": 0.8119178081100311,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019731139368410954,
+      "loss": 2.8139,
+      "step": 15922
+    },
+    {
+      "epoch": 0.8119688015661365,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001972942184277175,
+      "loss": 2.8167,
+      "step": 15923
+    },
+    {
+      "epoch": 0.8120197950222419,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0019727698932386567,
+      "loss": 2.8248,
+      "step": 15924
+    },
+    {
+      "epoch": 0.8120707884783473,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001972597063831817,
+      "loss": 2.8234,
+      "step": 15925
+    },
+    {
+      "epoch": 0.8121217819344527,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0019724236961632664,
+      "loss": 2.8159,
+      "step": 15926
+    },
+    {
+      "epoch": 0.8121727753905581,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019722497903399464,
+      "loss": 2.8111,
+      "step": 15927
+    },
+    {
+      "epoch": 0.8122237688466635,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.001972075346469131,
+      "loss": 2.8154,
+      "step": 15928
+    },
+    {
+      "epoch": 0.8122747623027688,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019719003646584254,
+      "loss": 2.8169,
+      "step": 15929
+    },
+    {
+      "epoch": 0.8123257557588742,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001971724845015768,
+      "loss": 2.8034,
+      "step": 15930
+    },
+    {
+      "epoch": 0.8123767492149796,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001971548787649428,
+      "loss": 2.7877,
+      "step": 15931
+    },
+    {
+      "epoch": 0.8124277426710851,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001971372192668006,
+      "loss": 2.8165,
+      "step": 15932
+    },
+    {
+      "epoch": 0.8124787361271905,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.001971195060180435,
+      "loss": 2.8553,
+      "step": 15933
+    },
+    {
+      "epoch": 0.8125297295832958,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.001971017390295979,
+      "loss": 2.8244,
+      "step": 15934
+    },
+    {
+      "epoch": 0.8125807230394012,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001970839183124234,
+      "loss": 2.8091,
+      "step": 15935
+    },
+    {
+      "epoch": 0.8126317164955066,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001970660438775127,
+      "loss": 2.8236,
+      "step": 15936
+    },
+    {
+      "epoch": 0.812682709951612,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001970481157358916,
+      "loss": 2.8305,
+      "step": 15937
+    },
+    {
+      "epoch": 0.8127337034077173,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.001970301338986192,
+      "loss": 2.8354,
+      "step": 15938
+    },
+    {
+      "epoch": 0.8127846968638228,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019701209837678746,
+      "loss": 2.8194,
+      "step": 15939
+    },
+    {
+      "epoch": 0.8128356903199282,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019699400918152168,
+      "loss": 2.8118,
+      "step": 15940
+    },
+    {
+      "epoch": 0.8128866837760336,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001969758663239802,
+      "loss": 2.835,
+      "step": 15941
+    },
+    {
+      "epoch": 0.8129376772321389,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019695766981535434,
+      "loss": 2.8151,
+      "step": 15942
+    },
+    {
+      "epoch": 0.8129886706882443,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001969394196668687,
+      "loss": 2.8331,
+      "step": 15943
+    },
+    {
+      "epoch": 0.8130396641443497,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0019692111588978096,
+      "loss": 2.7907,
+      "step": 15944
+    },
+    {
+      "epoch": 0.8130906576004551,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0019690275849538167,
+      "loss": 2.788,
+      "step": 15945
+    },
+    {
+      "epoch": 0.8131416510565606,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019688434749499466,
+      "loss": 2.8249,
+      "step": 15946
+    },
+    {
+      "epoch": 0.8131926445126659,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019686588289997676,
+      "loss": 2.8372,
+      "step": 15947
+    },
+    {
+      "epoch": 0.8132436379687713,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019684736472171784,
+      "loss": 2.8039,
+      "step": 15948
+    },
+    {
+      "epoch": 0.8132946314248767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.001968287929716409,
+      "loss": 2.8136,
+      "step": 15949
+    },
+    {
+      "epoch": 0.813345624880982,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019681016766120185,
+      "loss": 2.8094,
+      "step": 15950
+    },
+    {
+      "epoch": 0.8133966183370874,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0019679148880188974,
+      "loss": 2.8232,
+      "step": 15951
+    },
+    {
+      "epoch": 0.8134476117931928,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019677275640522665,
+      "loss": 2.804,
+      "step": 15952
+    },
+    {
+      "epoch": 0.8134986052492983,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019675397048276766,
+      "loss": 2.8641,
+      "step": 15953
+    },
+    {
+      "epoch": 0.8135495987054037,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001967351310461009,
+      "loss": 2.8264,
+      "step": 15954
+    },
+    {
+      "epoch": 0.813600592161509,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001967162381068474,
+      "loss": 2.8318,
+      "step": 15955
+    },
+    {
+      "epoch": 0.8136515856176144,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.001966972916766613,
+      "loss": 2.8024,
+      "step": 15956
+    },
+    {
+      "epoch": 0.8137025790737198,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.001966782917672298,
+      "loss": 2.831,
+      "step": 15957
+    },
+    {
+      "epoch": 0.8137535725298252,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001966592383902729,
+      "loss": 2.8186,
+      "step": 15958
+    },
+    {
+      "epoch": 0.8138045659859305,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001966401315575437,
+      "loss": 2.784,
+      "step": 15959
+    },
+    {
+      "epoch": 0.813855559442036,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0019662097128082825,
+      "loss": 2.8278,
+      "step": 15960
+    },
+    {
+      "epoch": 0.8139065528981414,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001966017575719456,
+      "loss": 2.8218,
+      "step": 15961
+    },
+    {
+      "epoch": 0.8139575463542468,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001965824904427477,
+      "loss": 2.8026,
+      "step": 15962
+    },
+    {
+      "epoch": 0.8140085398103521,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0019656316990511956,
+      "loss": 2.8249,
+      "step": 15963
+    },
+    {
+      "epoch": 0.8140595332664575,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0019654379597097895,
+      "loss": 2.8262,
+      "step": 15964
+    },
+    {
+      "epoch": 0.8141105267225629,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0019652436865227677,
+      "loss": 2.7839,
+      "step": 15965
+    },
+    {
+      "epoch": 0.8141615201786683,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001965048879609967,
+      "loss": 2.8107,
+      "step": 15966
+    },
+    {
+      "epoch": 0.8142125136347738,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019648535390915543,
+      "loss": 2.8073,
+      "step": 15967
+    },
+    {
+      "epoch": 0.8142635070908791,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019646576650880253,
+      "loss": 2.8401,
+      "step": 15968
+    },
+    {
+      "epoch": 0.8143145005469845,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001964461257720206,
+      "loss": 2.818,
+      "step": 15969
+    },
+    {
+      "epoch": 0.8143654940030899,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.001964264317109249,
+      "loss": 2.8099,
+      "step": 15970
+    },
+    {
+      "epoch": 0.8144164874591953,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001964066843376637,
+      "loss": 2.8275,
+      "step": 15971
+    },
+    {
+      "epoch": 0.8144674809153006,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019638688366441833,
+      "loss": 2.8315,
+      "step": 15972
+    },
+    {
+      "epoch": 0.814518474371406,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001963670297034028,
+      "loss": 2.8247,
+      "step": 15973
+    },
+    {
+      "epoch": 0.8145694678275115,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0019634712246686388,
+      "loss": 2.8042,
+      "step": 15974
+    },
+    {
+      "epoch": 0.8146204612836169,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019632716196708146,
+      "loss": 2.8,
+      "step": 15975
+    },
+    {
+      "epoch": 0.8146714547397222,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.001963071482163682,
+      "loss": 2.8361,
+      "step": 15976
+    },
+    {
+      "epoch": 0.8147224481958276,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019628708122706954,
+      "loss": 2.8061,
+      "step": 15977
+    },
+    {
+      "epoch": 0.814773441651933,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019626696101156384,
+      "loss": 2.8282,
+      "step": 15978
+    },
+    {
+      "epoch": 0.8148244351080384,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019624678758226226,
+      "loss": 2.8477,
+      "step": 15979
+    },
+    {
+      "epoch": 0.8148754285641437,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0019622656095160873,
+      "loss": 2.8247,
+      "step": 15980
+    },
+    {
+      "epoch": 0.8149264220202492,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.001962062811320801,
+      "loss": 2.8414,
+      "step": 15981
+    },
+    {
+      "epoch": 0.8149774154763546,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00196185948136186,
+      "loss": 2.8153,
+      "step": 15982
+    },
+    {
+      "epoch": 0.81502840893246,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001961655619764688,
+      "loss": 2.8526,
+      "step": 15983
+    },
+    {
+      "epoch": 0.8150794023885654,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0019614512266550375,
+      "loss": 2.8021,
+      "step": 15984
+    },
+    {
+      "epoch": 0.8151303958446707,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0019612463021589878,
+      "loss": 2.8142,
+      "step": 15985
+    },
+    {
+      "epoch": 0.8151813893007761,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0019610408464029476,
+      "loss": 2.8366,
+      "step": 15986
+    },
+    {
+      "epoch": 0.8152323827568815,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0019608348595136515,
+      "loss": 2.8294,
+      "step": 15987
+    },
+    {
+      "epoch": 0.815283376212987,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0019606283416181624,
+      "loss": 2.8185,
+      "step": 15988
+    },
+    {
+      "epoch": 0.8153343696690923,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0019604212928438715,
+      "loss": 2.8347,
+      "step": 15989
+    },
+    {
+      "epoch": 0.8153853631251977,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001960213713318497,
+      "loss": 2.8041,
+      "step": 15990
+    },
+    {
+      "epoch": 0.8154363565813031,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001960005603170084,
+      "loss": 2.8064,
+      "step": 15991
+    },
+    {
+      "epoch": 0.8154873500374085,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019597969625270058,
+      "loss": 2.8038,
+      "step": 15992
+    },
+    {
+      "epoch": 0.8155383434935138,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001959587791517961,
+      "loss": 2.8057,
+      "step": 15993
+    },
+    {
+      "epoch": 0.8155893369496192,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.001959378090271979,
+      "loss": 2.8033,
+      "step": 15994
+    },
+    {
+      "epoch": 0.8156403304057247,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.001959167858918412,
+      "loss": 2.8119,
+      "step": 15995
+    },
+    {
+      "epoch": 0.8156913238618301,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019589570975869423,
+      "loss": 2.8257,
+      "step": 15996
+    },
+    {
+      "epoch": 0.8157423173179355,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019587458064075783,
+      "loss": 2.8228,
+      "step": 15997
+    },
+    {
+      "epoch": 0.8157933107740408,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001958533985510654,
+      "loss": 2.8153,
+      "step": 15998
+    },
+    {
+      "epoch": 0.8158443042301462,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001958321635026832,
+      "loss": 2.819,
+      "step": 15999
+    },
+    {
+      "epoch": 0.8158952976862516,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001958108755087101,
+      "loss": 2.8205,
+      "step": 16000
+    },
+    {
+      "epoch": 0.815946291142357,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019578953458227757,
+      "loss": 2.8326,
+      "step": 16001
+    },
+    {
+      "epoch": 0.8159972845984624,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.001957681407365497,
+      "loss": 2.8167,
+      "step": 16002
+    },
+    {
+      "epoch": 0.8160482780545678,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019574669398472336,
+      "loss": 2.8064,
+      "step": 16003
+    },
+    {
+      "epoch": 0.8160992715106732,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019572519434002795,
+      "loss": 2.7993,
+      "step": 16004
+    },
+    {
+      "epoch": 0.8161502649667786,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.001957036418157256,
+      "loss": 2.8284,
+      "step": 16005
+    },
+    {
+      "epoch": 0.8162012584228839,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001956820364251109,
+      "loss": 2.8356,
+      "step": 16006
+    },
+    {
+      "epoch": 0.8162522518789893,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019566037818151117,
+      "loss": 2.8271,
+      "step": 16007
+    },
+    {
+      "epoch": 0.8163032453350947,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001956386670982863,
+      "loss": 2.8277,
+      "step": 16008
+    },
+    {
+      "epoch": 0.8163542387912002,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0019561690318882887,
+      "loss": 2.8171,
+      "step": 16009
+    },
+    {
+      "epoch": 0.8164052322473055,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0019559508646656383,
+      "loss": 2.8317,
+      "step": 16010
+    },
+    {
+      "epoch": 0.8164562257034109,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019557321694494887,
+      "loss": 2.8215,
+      "step": 16011
+    },
+    {
+      "epoch": 0.8165072191595163,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001955512946374742,
+      "loss": 2.83,
+      "step": 16012
+    },
+    {
+      "epoch": 0.8165582126156217,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0019552931955766264,
+      "loss": 2.824,
+      "step": 16013
+    },
+    {
+      "epoch": 0.816609206071727,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0019550729171906945,
+      "loss": 2.8231,
+      "step": 16014
+    },
+    {
+      "epoch": 0.8166601995278324,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001954852111352826,
+      "loss": 2.8141,
+      "step": 16015
+    },
+    {
+      "epoch": 0.8167111929839379,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019546307781992244,
+      "loss": 2.8282,
+      "step": 16016
+    },
+    {
+      "epoch": 0.8167621864400433,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001954408917866419,
+      "loss": 2.8075,
+      "step": 16017
+    },
+    {
+      "epoch": 0.8168131798961487,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001954186530491265,
+      "loss": 2.8385,
+      "step": 16018
+    },
+    {
+      "epoch": 0.816864173352254,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.001953963616210942,
+      "loss": 2.8226,
+      "step": 16019
+    },
+    {
+      "epoch": 0.8169151668083594,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001953740175162954,
+      "loss": 2.7837,
+      "step": 16020
+    },
+    {
+      "epoch": 0.8169661602644648,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019535162074851317,
+      "loss": 2.7961,
+      "step": 16021
+    },
+    {
+      "epoch": 0.8170171537205702,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0019532917133156286,
+      "loss": 2.8359,
+      "step": 16022
+    },
+    {
+      "epoch": 0.8170681471766756,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001953066692792925,
+      "loss": 2.8294,
+      "step": 16023
+    },
+    {
+      "epoch": 0.817119140632781,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.001952841146055824,
+      "loss": 2.8256,
+      "step": 16024
+    },
+    {
+      "epoch": 0.8171701340888864,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019526150732434547,
+      "loss": 2.8244,
+      "step": 16025
+    },
+    {
+      "epoch": 0.8172211275449918,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0019523884744952701,
+      "loss": 2.7967,
+      "step": 16026
+    },
+    {
+      "epoch": 0.8172721210010971,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001952161349951047,
+      "loss": 2.7883,
+      "step": 16027
+    },
+    {
+      "epoch": 0.8173231144572025,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0019519336997508884,
+      "loss": 2.8147,
+      "step": 16028
+    },
+    {
+      "epoch": 0.8173741079133079,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019517055240352195,
+      "loss": 2.7948,
+      "step": 16029
+    },
+    {
+      "epoch": 0.8174251013694134,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019514768229447908,
+      "loss": 2.804,
+      "step": 16030
+    },
+    {
+      "epoch": 0.8174760948255188,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019512475966206768,
+      "loss": 2.825,
+      "step": 16031
+    },
+    {
+      "epoch": 0.8175270882816241,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0019510178452042755,
+      "loss": 2.7958,
+      "step": 16032
+    },
+    {
+      "epoch": 0.8175780817377295,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019507875688373093,
+      "loss": 2.8166,
+      "step": 16033
+    },
+    {
+      "epoch": 0.8176290751938349,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019505567676618243,
+      "loss": 2.8213,
+      "step": 16034
+    },
+    {
+      "epoch": 0.8176800686499403,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0019503254418201904,
+      "loss": 2.8028,
+      "step": 16035
+    },
+    {
+      "epoch": 0.8177310621060456,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0019500935914551006,
+      "loss": 2.7999,
+      "step": 16036
+    },
+    {
+      "epoch": 0.8177820555621511,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0019498612167095717,
+      "loss": 2.8044,
+      "step": 16037
+    },
+    {
+      "epoch": 0.8178330490182565,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001949628317726945,
+      "loss": 2.8165,
+      "step": 16038
+    },
+    {
+      "epoch": 0.8178840424743619,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019493948946508832,
+      "loss": 2.8284,
+      "step": 16039
+    },
+    {
+      "epoch": 0.8179350359304672,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019491609476253741,
+      "loss": 2.8226,
+      "step": 16040
+    },
+    {
+      "epoch": 0.8179860293865726,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019489264767947282,
+      "loss": 2.7961,
+      "step": 16041
+    },
+    {
+      "epoch": 0.818037022842678,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0019486914823035782,
+      "loss": 2.8156,
+      "step": 16042
+    },
+    {
+      "epoch": 0.8180880162987834,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0019484559642968806,
+      "loss": 2.7946,
+      "step": 16043
+    },
+    {
+      "epoch": 0.8181390097548888,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0019482199229199151,
+      "loss": 2.8344,
+      "step": 16044
+    },
+    {
+      "epoch": 0.8181900032109942,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019479833583182838,
+      "loss": 2.8197,
+      "step": 16045
+    },
+    {
+      "epoch": 0.8182409966670996,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0019477462706379114,
+      "loss": 2.8561,
+      "step": 16046
+    },
+    {
+      "epoch": 0.818291990123205,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001947508660025046,
+      "loss": 2.7875,
+      "step": 16047
+    },
+    {
+      "epoch": 0.8183429835793103,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019472705266262574,
+      "loss": 2.8083,
+      "step": 16048
+    },
+    {
+      "epoch": 0.8183939770354157,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0019470318705884378,
+      "loss": 2.8121,
+      "step": 16049
+    },
+    {
+      "epoch": 0.8184449704915211,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001946792692058803,
+      "loss": 2.8339,
+      "step": 16050
+    },
+    {
+      "epoch": 0.8184959639476266,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00194655299118489,
+      "loss": 2.7934,
+      "step": 16051
+    },
+    {
+      "epoch": 0.818546957403732,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0019463127681145585,
+      "loss": 2.8304,
+      "step": 16052
+    },
+    {
+      "epoch": 0.8185979508598373,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0019460720229959895,
+      "loss": 2.8102,
+      "step": 16053
+    },
+    {
+      "epoch": 0.8186489443159427,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019458307559776877,
+      "loss": 2.8105,
+      "step": 16054
+    },
+    {
+      "epoch": 0.8186999377720481,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001945588967208478,
+      "loss": 2.8177,
+      "step": 16055
+    },
+    {
+      "epoch": 0.8187509312281535,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019453466568375084,
+      "loss": 2.8307,
+      "step": 16056
+    },
+    {
+      "epoch": 0.8188019246842588,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0019451038250142476,
+      "loss": 2.7945,
+      "step": 16057
+    },
+    {
+      "epoch": 0.8188529181403643,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0019448604718884867,
+      "loss": 2.8261,
+      "step": 16058
+    },
+    {
+      "epoch": 0.8189039115964697,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001944616597610338,
+      "loss": 2.8147,
+      "step": 16059
+    },
+    {
+      "epoch": 0.8189549050525751,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0019443722023302359,
+      "loss": 2.8367,
+      "step": 16060
+    },
+    {
+      "epoch": 0.8190058985086804,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019441272861989347,
+      "loss": 2.8178,
+      "step": 16061
+    },
+    {
+      "epoch": 0.8190568919647858,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001943881849367512,
+      "loss": 2.8169,
+      "step": 16062
+    },
+    {
+      "epoch": 0.8191078854208912,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0019436358919873652,
+      "loss": 2.8404,
+      "step": 16063
+    },
+    {
+      "epoch": 0.8191588788769966,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.001943389414210213,
+      "loss": 2.8083,
+      "step": 16064
+    },
+    {
+      "epoch": 0.819209872333102,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0019431424161880952,
+      "loss": 2.8096,
+      "step": 16065
+    },
+    {
+      "epoch": 0.8192608657892074,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019428948980733732,
+      "loss": 2.8025,
+      "step": 16066
+    },
+    {
+      "epoch": 0.8193118592453128,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001942646860018728,
+      "loss": 2.8286,
+      "step": 16067
+    },
+    {
+      "epoch": 0.8193628527014182,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019423983021771621,
+      "loss": 2.8199,
+      "step": 16068
+    },
+    {
+      "epoch": 0.8194138461575236,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019421492247019985,
+      "loss": 2.8133,
+      "step": 16069
+    },
+    {
+      "epoch": 0.8194648396136289,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0019418996277468807,
+      "loss": 2.8123,
+      "step": 16070
+    },
+    {
+      "epoch": 0.8195158330697343,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019416495114657727,
+      "loss": 2.8014,
+      "step": 16071
+    },
+    {
+      "epoch": 0.8195668265258398,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0019413988760129583,
+      "loss": 2.8035,
+      "step": 16072
+    },
+    {
+      "epoch": 0.8196178199819452,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001941147721543043,
+      "loss": 2.8028,
+      "step": 16073
+    },
+    {
+      "epoch": 0.8196688134380505,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001940896048210951,
+      "loss": 2.8073,
+      "step": 16074
+    },
+    {
+      "epoch": 0.8197198068941559,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019406438561719267,
+      "loss": 2.812,
+      "step": 16075
+    },
+    {
+      "epoch": 0.8197708003502613,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0019403911455815352,
+      "loss": 2.8023,
+      "step": 16076
+    },
+    {
+      "epoch": 0.8198217938063667,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019401379165956613,
+      "loss": 2.8207,
+      "step": 16077
+    },
+    {
+      "epoch": 0.819872787262472,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0019398841693705087,
+      "loss": 2.8053,
+      "step": 16078
+    },
+    {
+      "epoch": 0.8199237807185775,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019396299040626021,
+      "loss": 2.8296,
+      "step": 16079
+    },
+    {
+      "epoch": 0.8199747741746829,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0019393751208287846,
+      "loss": 2.8306,
+      "step": 16080
+    },
+    {
+      "epoch": 0.8200257676307883,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00193911981982622,
+      "loss": 2.8174,
+      "step": 16081
+    },
+    {
+      "epoch": 0.8200767610868936,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019388640012123898,
+      "loss": 2.839,
+      "step": 16082
+    },
+    {
+      "epoch": 0.820127754542999,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0019386076651450965,
+      "loss": 2.829,
+      "step": 16083
+    },
+    {
+      "epoch": 0.8201787479991044,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001938350811782461,
+      "loss": 2.8256,
+      "step": 16084
+    },
+    {
+      "epoch": 0.8202297414552098,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0019380934412829233,
+      "loss": 2.8073,
+      "step": 16085
+    },
+    {
+      "epoch": 0.8202807349113153,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001937835553805242,
+      "loss": 2.822,
+      "step": 16086
+    },
+    {
+      "epoch": 0.8203317283674206,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019375771495084959,
+      "loss": 2.8176,
+      "step": 16087
+    },
+    {
+      "epoch": 0.820382721823526,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019373182285520813,
+      "loss": 2.8039,
+      "step": 16088
+    },
+    {
+      "epoch": 0.8204337152796314,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019370587910957132,
+      "loss": 2.7852,
+      "step": 16089
+    },
+    {
+      "epoch": 0.8204847087357368,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0019367988372994264,
+      "loss": 2.8088,
+      "step": 16090
+    },
+    {
+      "epoch": 0.8205357021918421,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001936538367323573,
+      "loss": 2.8156,
+      "step": 16091
+    },
+    {
+      "epoch": 0.8205866956479475,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019362773813288243,
+      "loss": 2.8099,
+      "step": 16092
+    },
+    {
+      "epoch": 0.820637689104053,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019360158794761693,
+      "loss": 2.8378,
+      "step": 16093
+    },
+    {
+      "epoch": 0.8206886825601584,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019357538619269158,
+      "loss": 2.7883,
+      "step": 16094
+    },
+    {
+      "epoch": 0.8207396760162637,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019354913288426889,
+      "loss": 2.8144,
+      "step": 16095
+    },
+    {
+      "epoch": 0.8207906694723691,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0019352282803854327,
+      "loss": 2.8172,
+      "step": 16096
+    },
+    {
+      "epoch": 0.8208416629284745,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0019349647167174083,
+      "loss": 2.8026,
+      "step": 16097
+    },
+    {
+      "epoch": 0.8208926563845799,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019347006380011954,
+      "loss": 2.8133,
+      "step": 16098
+    },
+    {
+      "epoch": 0.8209436498406852,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019344360443996904,
+      "loss": 2.8243,
+      "step": 16099
+    },
+    {
+      "epoch": 0.8209946432967907,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0019341709360761086,
+      "loss": 2.8285,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8210456367528961,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019339053131939823,
+      "loss": 2.7858,
+      "step": 16101
+    },
+    {
+      "epoch": 0.8210966302090015,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00193363917591716,
+      "loss": 2.8113,
+      "step": 16102
+    },
+    {
+      "epoch": 0.8211476236651069,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019333725244098095,
+      "loss": 2.8019,
+      "step": 16103
+    },
+    {
+      "epoch": 0.8211986171212122,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019331053588364144,
+      "loss": 2.8095,
+      "step": 16104
+    },
+    {
+      "epoch": 0.8212496105773176,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.001932837679361776,
+      "loss": 2.808,
+      "step": 16105
+    },
+    {
+      "epoch": 0.821300604033423,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0019325694861510121,
+      "loss": 2.7961,
+      "step": 16106
+    },
+    {
+      "epoch": 0.8213515974895285,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0019323007793695586,
+      "loss": 2.8222,
+      "step": 16107
+    },
+    {
+      "epoch": 0.8214025909456338,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019320315591831666,
+      "loss": 2.8075,
+      "step": 16108
+    },
+    {
+      "epoch": 0.8214535844017392,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0019317618257579047,
+      "loss": 2.8371,
+      "step": 16109
+    },
+    {
+      "epoch": 0.8215045778578446,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001931491579260158,
+      "loss": 2.8374,
+      "step": 16110
+    },
+    {
+      "epoch": 0.82155557131395,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019312208198566285,
+      "loss": 2.8136,
+      "step": 16111
+    },
+    {
+      "epoch": 0.8216065647700553,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.001930949547714334,
+      "loss": 2.804,
+      "step": 16112
+    },
+    {
+      "epoch": 0.8216575582261607,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0019306777630006089,
+      "loss": 2.7988,
+      "step": 16113
+    },
+    {
+      "epoch": 0.8217085516822662,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019304054658831034,
+      "loss": 2.8181,
+      "step": 16114
+    },
+    {
+      "epoch": 0.8217595451383716,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.001930132656529784,
+      "loss": 2.7782,
+      "step": 16115
+    },
+    {
+      "epoch": 0.821810538594477,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0019298593351089338,
+      "loss": 2.8186,
+      "step": 16116
+    },
+    {
+      "epoch": 0.8218615320505823,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0019295855017891505,
+      "loss": 2.8261,
+      "step": 16117
+    },
+    {
+      "epoch": 0.8219125255066877,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019293111567393487,
+      "loss": 2.8401,
+      "step": 16118
+    },
+    {
+      "epoch": 0.8219635189627931,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019290363001287579,
+      "loss": 2.8198,
+      "step": 16119
+    },
+    {
+      "epoch": 0.8220145124188984,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019287609321269237,
+      "loss": 2.8016,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8220655058750039,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019284850529037068,
+      "loss": 2.8215,
+      "step": 16121
+    },
+    {
+      "epoch": 0.8221164993311093,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019282086626292833,
+      "loss": 2.7806,
+      "step": 16122
+    },
+    {
+      "epoch": 0.8221674927872147,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001927931761474145,
+      "loss": 2.7879,
+      "step": 16123
+    },
+    {
+      "epoch": 0.8222184862433201,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019276543496090977,
+      "loss": 2.7947,
+      "step": 16124
+    },
+    {
+      "epoch": 0.8222694796994254,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0019273764272052637,
+      "loss": 2.7921,
+      "step": 16125
+    },
+    {
+      "epoch": 0.8223204731555308,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019270979944340791,
+      "loss": 2.8131,
+      "step": 16126
+    },
+    {
+      "epoch": 0.8223714666116362,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0019268190514672957,
+      "loss": 2.8116,
+      "step": 16127
+    },
+    {
+      "epoch": 0.8224224600677417,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0019265395984769786,
+      "loss": 2.8322,
+      "step": 16128
+    },
+    {
+      "epoch": 0.822473453523847,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019262596356355095,
+      "loss": 2.7972,
+      "step": 16129
+    },
+    {
+      "epoch": 0.8225244469799524,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0019259791631155829,
+      "loss": 2.8208,
+      "step": 16130
+    },
+    {
+      "epoch": 0.8225754404360578,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0019256981810902088,
+      "loss": 2.809,
+      "step": 16131
+    },
+    {
+      "epoch": 0.8226264338921632,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019254166897327103,
+      "loss": 2.8104,
+      "step": 16132
+    },
+    {
+      "epoch": 0.8226774273482685,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019251346892167262,
+      "loss": 2.8374,
+      "step": 16133
+    },
+    {
+      "epoch": 0.8227284208043739,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019248521797162082,
+      "loss": 2.8133,
+      "step": 16134
+    },
+    {
+      "epoch": 0.8227794142604794,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019245691614054226,
+      "loss": 2.8352,
+      "step": 16135
+    },
+    {
+      "epoch": 0.8228304077165848,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0019242856344589485,
+      "loss": 2.8187,
+      "step": 16136
+    },
+    {
+      "epoch": 0.8228814011726902,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0019240015990516807,
+      "loss": 2.8072,
+      "step": 16137
+    },
+    {
+      "epoch": 0.8229323946287955,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.001923717055358826,
+      "loss": 2.7923,
+      "step": 16138
+    },
+    {
+      "epoch": 0.8229833880849009,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0019234320035559048,
+      "loss": 2.8027,
+      "step": 16139
+    },
+    {
+      "epoch": 0.8230343815410063,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0019231464438187523,
+      "loss": 2.8241,
+      "step": 16140
+    },
+    {
+      "epoch": 0.8230853749971117,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019228603763235155,
+      "loss": 2.8221,
+      "step": 16141
+    },
+    {
+      "epoch": 0.8231363684532171,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019225738012466549,
+      "loss": 2.7717,
+      "step": 16142
+    },
+    {
+      "epoch": 0.8231873619093225,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019222867187649454,
+      "loss": 2.8011,
+      "step": 16143
+    },
+    {
+      "epoch": 0.8232383553654279,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001921999129055473,
+      "loss": 2.8216,
+      "step": 16144
+    },
+    {
+      "epoch": 0.8232893488215333,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001921711032295638,
+      "loss": 2.8183,
+      "step": 16145
+    },
+    {
+      "epoch": 0.8233403422776386,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0019214224286631526,
+      "loss": 2.8138,
+      "step": 16146
+    },
+    {
+      "epoch": 0.823391335733744,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001921133318336042,
+      "loss": 2.8039,
+      "step": 16147
+    },
+    {
+      "epoch": 0.8234423291898494,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019208437014926446,
+      "loss": 2.7822,
+      "step": 16148
+    },
+    {
+      "epoch": 0.8234933226459549,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0019205535783116097,
+      "loss": 2.8301,
+      "step": 16149
+    },
+    {
+      "epoch": 0.8235443161020602,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.001920262948971901,
+      "loss": 2.8128,
+      "step": 16150
+    },
+    {
+      "epoch": 0.8235953095581656,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0019199718136527922,
+      "loss": 2.8073,
+      "step": 16151
+    },
+    {
+      "epoch": 0.823646303014271,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0019196801725338702,
+      "loss": 2.8137,
+      "step": 16152
+    },
+    {
+      "epoch": 0.8236972964703764,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001919388025795035,
+      "loss": 2.8189,
+      "step": 16153
+    },
+    {
+      "epoch": 0.8237482899264817,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0019190953736164963,
+      "loss": 2.7967,
+      "step": 16154
+    },
+    {
+      "epoch": 0.8237992833825871,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.001918802216178777,
+      "loss": 2.8126,
+      "step": 16155
+    },
+    {
+      "epoch": 0.8238502768386926,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019185085536627114,
+      "loss": 2.7971,
+      "step": 16156
+    },
+    {
+      "epoch": 0.823901270294798,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019182143862494455,
+      "loss": 2.7881,
+      "step": 16157
+    },
+    {
+      "epoch": 0.8239522637509034,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019179197141204363,
+      "loss": 2.79,
+      "step": 16158
+    },
+    {
+      "epoch": 0.8240032572070087,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019176245374574522,
+      "loss": 2.7992,
+      "step": 16159
+    },
+    {
+      "epoch": 0.8240542506631141,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019173288564425734,
+      "loss": 2.8219,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8241052441192195,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019170326712581902,
+      "loss": 2.8319,
+      "step": 16161
+    },
+    {
+      "epoch": 0.8241562375753249,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019167359820870052,
+      "loss": 2.7848,
+      "step": 16162
+    },
+    {
+      "epoch": 0.8242072310314303,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019164387891120313,
+      "loss": 2.8048,
+      "step": 16163
+    },
+    {
+      "epoch": 0.8242582244875357,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001916141092516591,
+      "loss": 2.8322,
+      "step": 16164
+    },
+    {
+      "epoch": 0.8243092179436411,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019158428924843192,
+      "loss": 2.8142,
+      "step": 16165
+    },
+    {
+      "epoch": 0.8243602113997465,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0019155441891991608,
+      "loss": 2.8103,
+      "step": 16166
+    },
+    {
+      "epoch": 0.8244112048558518,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001915244982845371,
+      "loss": 2.8025,
+      "step": 16167
+    },
+    {
+      "epoch": 0.8244621983119572,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001914945273607515,
+      "loss": 2.809,
+      "step": 16168
+    },
+    {
+      "epoch": 0.8245131917680626,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019146450616704684,
+      "loss": 2.834,
+      "step": 16169
+    },
+    {
+      "epoch": 0.8245641852241681,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019143443472194178,
+      "loss": 2.8104,
+      "step": 16170
+    },
+    {
+      "epoch": 0.8246151786802735,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001914043130439858,
+      "loss": 2.8421,
+      "step": 16171
+    },
+    {
+      "epoch": 0.8246661721363788,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019137414115175952,
+      "loss": 2.8023,
+      "step": 16172
+    },
+    {
+      "epoch": 0.8247171655924842,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0019134391906387444,
+      "loss": 2.8116,
+      "step": 16173
+    },
+    {
+      "epoch": 0.8247681590485896,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0019131364679897313,
+      "loss": 2.8101,
+      "step": 16174
+    },
+    {
+      "epoch": 0.824819152504695,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019128332437572902,
+      "loss": 2.8193,
+      "step": 16175
+    },
+    {
+      "epoch": 0.8248701459608003,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0019125295181284644,
+      "loss": 2.8037,
+      "step": 16176
+    },
+    {
+      "epoch": 0.8249211394169058,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0019122252912906078,
+      "loss": 2.8043,
+      "step": 16177
+    },
+    {
+      "epoch": 0.8249721328730112,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0019119205634313826,
+      "loss": 2.8282,
+      "step": 16178
+    },
+    {
+      "epoch": 0.8250231263291166,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0019116153347387601,
+      "loss": 2.8065,
+      "step": 16179
+    },
+    {
+      "epoch": 0.8250741197852219,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0019113096054010211,
+      "loss": 2.8142,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8251251132413273,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019110033756067546,
+      "loss": 2.8262,
+      "step": 16181
+    },
+    {
+      "epoch": 0.8251761066974327,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019106966455448582,
+      "loss": 2.7693,
+      "step": 16182
+    },
+    {
+      "epoch": 0.8252271001535381,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019103894154045389,
+      "loss": 2.8044,
+      "step": 16183
+    },
+    {
+      "epoch": 0.8252780936096435,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0019100816853753115,
+      "loss": 2.8259,
+      "step": 16184
+    },
+    {
+      "epoch": 0.8253290870657489,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0019097734556469995,
+      "loss": 2.7879,
+      "step": 16185
+    },
+    {
+      "epoch": 0.8253800805218543,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001909464726409734,
+      "loss": 2.8083,
+      "step": 16186
+    },
+    {
+      "epoch": 0.8254310739779597,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019091554978539554,
+      "loss": 2.7949,
+      "step": 16187
+    },
+    {
+      "epoch": 0.825482067434065,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019088457701704107,
+      "loss": 2.7977,
+      "step": 16188
+    },
+    {
+      "epoch": 0.8255330608901704,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019085355435501557,
+      "loss": 2.799,
+      "step": 16189
+    },
+    {
+      "epoch": 0.8255840543462758,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001908224818184554,
+      "loss": 2.8134,
+      "step": 16190
+    },
+    {
+      "epoch": 0.8256350478023813,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019079135942652768,
+      "loss": 2.8126,
+      "step": 16191
+    },
+    {
+      "epoch": 0.8256860412584867,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0019076018719843023,
+      "loss": 2.7833,
+      "step": 16192
+    },
+    {
+      "epoch": 0.825737034714592,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019072896515339162,
+      "loss": 2.824,
+      "step": 16193
+    },
+    {
+      "epoch": 0.8257880281706974,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019069769331067125,
+      "loss": 2.7972,
+      "step": 16194
+    },
+    {
+      "epoch": 0.8258390216268028,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.001906663716895591,
+      "loss": 2.8187,
+      "step": 16195
+    },
+    {
+      "epoch": 0.8258900150829082,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0019063500030937592,
+      "loss": 2.8125,
+      "step": 16196
+    },
+    {
+      "epoch": 0.8259410085390135,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019060357918947322,
+      "loss": 2.788,
+      "step": 16197
+    },
+    {
+      "epoch": 0.825992001995119,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0019057210834923306,
+      "loss": 2.8412,
+      "step": 16198
+    },
+    {
+      "epoch": 0.8260429954512244,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0019054058780806824,
+      "loss": 2.7967,
+      "step": 16199
+    },
+    {
+      "epoch": 0.8260939889073298,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0019050901758542226,
+      "loss": 2.7717,
+      "step": 16200
+    },
+    {
+      "epoch": 0.8261449823634351,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019047739770076918,
+      "loss": 2.8209,
+      "step": 16201
+    },
+    {
+      "epoch": 0.8261959758195405,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0019044572817361375,
+      "loss": 2.8094,
+      "step": 16202
+    },
+    {
+      "epoch": 0.8262469692756459,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001904140090234913,
+      "loss": 2.8154,
+      "step": 16203
+    },
+    {
+      "epoch": 0.8262979627317513,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019038224026996782,
+      "loss": 2.8062,
+      "step": 16204
+    },
+    {
+      "epoch": 0.8263489561878568,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019035042193263987,
+      "loss": 2.8022,
+      "step": 16205
+    },
+    {
+      "epoch": 0.8263999496439621,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001903185540311346,
+      "loss": 2.7762,
+      "step": 16206
+    },
+    {
+      "epoch": 0.8264509431000675,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019028663658510975,
+      "loss": 2.8131,
+      "step": 16207
+    },
+    {
+      "epoch": 0.8265019365561729,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019025466961425355,
+      "loss": 2.8102,
+      "step": 16208
+    },
+    {
+      "epoch": 0.8265529300122783,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019022265313828489,
+      "loss": 2.8038,
+      "step": 16209
+    },
+    {
+      "epoch": 0.8266039234683836,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001901905871769531,
+      "loss": 2.8173,
+      "step": 16210
+    },
+    {
+      "epoch": 0.826654916924489,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0019015847175003812,
+      "loss": 2.8088,
+      "step": 16211
+    },
+    {
+      "epoch": 0.8267059103805945,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001901263068773503,
+      "loss": 2.7986,
+      "step": 16212
+    },
+    {
+      "epoch": 0.8267569038366999,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0019009409257873062,
+      "loss": 2.8251,
+      "step": 16213
+    },
+    {
+      "epoch": 0.8268078972928052,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019006182887405043,
+      "loss": 2.8188,
+      "step": 16214
+    },
+    {
+      "epoch": 0.8268588907489106,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001900295157832116,
+      "loss": 2.7957,
+      "step": 16215
+    },
+    {
+      "epoch": 0.826909884205016,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0018999715332614648,
+      "loss": 2.8022,
+      "step": 16216
+    },
+    {
+      "epoch": 0.8269608776611214,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018996474152281786,
+      "loss": 2.7915,
+      "step": 16217
+    },
+    {
+      "epoch": 0.8270118711172267,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00189932280393219,
+      "loss": 2.8164,
+      "step": 16218
+    },
+    {
+      "epoch": 0.8270628645733322,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0018989976995737352,
+      "loss": 2.8222,
+      "step": 16219
+    },
+    {
+      "epoch": 0.8271138580294376,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001898672102353355,
+      "loss": 2.8235,
+      "step": 16220
+    },
+    {
+      "epoch": 0.827164851485543,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.001898346012471894,
+      "loss": 2.8188,
+      "step": 16221
+    },
+    {
+      "epoch": 0.8272158449416483,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001898019430130501,
+      "loss": 2.806,
+      "step": 16222
+    },
+    {
+      "epoch": 0.8272668383977537,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0018976923555306282,
+      "loss": 2.8167,
+      "step": 16223
+    },
+    {
+      "epoch": 0.8273178318538591,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018973647888740315,
+      "loss": 2.8201,
+      "step": 16224
+    },
+    {
+      "epoch": 0.8273688253099645,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.001897036730362771,
+      "loss": 2.8012,
+      "step": 16225
+    },
+    {
+      "epoch": 0.82741981876607,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0018967081801992097,
+      "loss": 2.821,
+      "step": 16226
+    },
+    {
+      "epoch": 0.8274708122221753,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018963791385860134,
+      "loss": 2.8113,
+      "step": 16227
+    },
+    {
+      "epoch": 0.8275218056782807,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0018960496057261517,
+      "loss": 2.8184,
+      "step": 16228
+    },
+    {
+      "epoch": 0.8275727991343861,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0018957195818228964,
+      "loss": 2.8237,
+      "step": 16229
+    },
+    {
+      "epoch": 0.8276237925904915,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018953890670798243,
+      "loss": 2.8131,
+      "step": 16230
+    },
+    {
+      "epoch": 0.8276747860465968,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001895058061700812,
+      "loss": 2.8244,
+      "step": 16231
+    },
+    {
+      "epoch": 0.8277257795027022,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.001894726565890041,
+      "loss": 2.7809,
+      "step": 16232
+    },
+    {
+      "epoch": 0.8277767729588077,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0018943945798519943,
+      "loss": 2.8245,
+      "step": 16233
+    },
+    {
+      "epoch": 0.8278277664149131,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001894062103791458,
+      "loss": 2.8075,
+      "step": 16234
+    },
+    {
+      "epoch": 0.8278787598710184,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0018937291379135196,
+      "loss": 2.7886,
+      "step": 16235
+    },
+    {
+      "epoch": 0.8279297533271238,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018933956824235692,
+      "loss": 2.838,
+      "step": 16236
+    },
+    {
+      "epoch": 0.8279807467832292,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0018930617375272991,
+      "loss": 2.7977,
+      "step": 16237
+    },
+    {
+      "epoch": 0.8280317402393346,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0018927273034307034,
+      "loss": 2.8081,
+      "step": 16238
+    },
+    {
+      "epoch": 0.8280827336954399,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0018923923803400773,
+      "loss": 2.8143,
+      "step": 16239
+    },
+    {
+      "epoch": 0.8281337271515454,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0018920569684620187,
+      "loss": 2.7828,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8281847206076508,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0018917210680034267,
+      "loss": 2.8141,
+      "step": 16241
+    },
+    {
+      "epoch": 0.8282357140637562,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018913846791715006,
+      "loss": 2.8212,
+      "step": 16242
+    },
+    {
+      "epoch": 0.8282867075198616,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0018910478021737429,
+      "loss": 2.8096,
+      "step": 16243
+    },
+    {
+      "epoch": 0.8283377009759669,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001890710437217956,
+      "loss": 2.7943,
+      "step": 16244
+    },
+    {
+      "epoch": 0.8283886944320723,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0018903725845122433,
+      "loss": 2.7848,
+      "step": 16245
+    },
+    {
+      "epoch": 0.8284396878881777,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0018900342442650095,
+      "loss": 2.8097,
+      "step": 16246
+    },
+    {
+      "epoch": 0.8284906813442832,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00188969541668496,
+      "loss": 2.8348,
+      "step": 16247
+    },
+    {
+      "epoch": 0.8285416748003885,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0018893561019811005,
+      "loss": 2.8039,
+      "step": 16248
+    },
+    {
+      "epoch": 0.8285926682564939,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018890163003627374,
+      "loss": 2.8114,
+      "step": 16249
+    },
+    {
+      "epoch": 0.8286436617125993,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018886760120394772,
+      "loss": 2.8194,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8286946551687047,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018883352372212276,
+      "loss": 2.7929,
+      "step": 16251
+    },
+    {
+      "epoch": 0.82874564862481,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018879939761181944,
+      "loss": 2.8017,
+      "step": 16252
+    },
+    {
+      "epoch": 0.8287966420809154,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0018876522289408854,
+      "loss": 2.8128,
+      "step": 16253
+    },
+    {
+      "epoch": 0.8288476355370209,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0018873099959001076,
+      "loss": 2.8357,
+      "step": 16254
+    },
+    {
+      "epoch": 0.8288986289931263,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001886967277206967,
+      "loss": 2.8175,
+      "step": 16255
+    },
+    {
+      "epoch": 0.8289496224492316,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0018866240730728702,
+      "loss": 2.7905,
+      "step": 16256
+    },
+    {
+      "epoch": 0.829000615905337,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0018862803837095222,
+      "loss": 2.8095,
+      "step": 16257
+    },
+    {
+      "epoch": 0.8290516093614424,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.001885936209328928,
+      "loss": 2.8218,
+      "step": 16258
+    },
+    {
+      "epoch": 0.8291026028175478,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018855915501433925,
+      "loss": 2.8048,
+      "step": 16259
+    },
+    {
+      "epoch": 0.8291535962736531,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0018852464063655176,
+      "loss": 2.7943,
+      "step": 16260
+    },
+    {
+      "epoch": 0.8292045897297586,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0018849007782082057,
+      "loss": 2.7837,
+      "step": 16261
+    },
+    {
+      "epoch": 0.829255583185864,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0018845546658846583,
+      "loss": 2.8201,
+      "step": 16262
+    },
+    {
+      "epoch": 0.8293065766419694,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0018842080696083736,
+      "loss": 2.7974,
+      "step": 16263
+    },
+    {
+      "epoch": 0.8293575700980748,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001883860989593151,
+      "loss": 2.7844,
+      "step": 16264
+    },
+    {
+      "epoch": 0.8294085635541801,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001883513426053086,
+      "loss": 2.7967,
+      "step": 16265
+    },
+    {
+      "epoch": 0.8294595570102855,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0018831653792025732,
+      "loss": 2.8193,
+      "step": 16266
+    },
+    {
+      "epoch": 0.8295105504663909,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0018828168492563057,
+      "loss": 2.8117,
+      "step": 16267
+    },
+    {
+      "epoch": 0.8295615439224964,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0018824678364292744,
+      "loss": 2.7956,
+      "step": 16268
+    },
+    {
+      "epoch": 0.8296125373786017,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0018821183409367676,
+      "loss": 2.8241,
+      "step": 16269
+    },
+    {
+      "epoch": 0.8296635308347071,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0018817683629943722,
+      "loss": 2.8081,
+      "step": 16270
+    },
+    {
+      "epoch": 0.8297145242908125,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018814179028179717,
+      "loss": 2.7961,
+      "step": 16271
+    },
+    {
+      "epoch": 0.8297655177469179,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018810669606237477,
+      "loss": 2.8344,
+      "step": 16272
+    },
+    {
+      "epoch": 0.8298165112030232,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0018807155366281788,
+      "loss": 2.8082,
+      "step": 16273
+    },
+    {
+      "epoch": 0.8298675046591286,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001880363631048041,
+      "loss": 2.8021,
+      "step": 16274
+    },
+    {
+      "epoch": 0.8299184981152341,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018800112441004076,
+      "loss": 2.7884,
+      "step": 16275
+    },
+    {
+      "epoch": 0.8299694915713395,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001879658376002649,
+      "loss": 2.8041,
+      "step": 16276
+    },
+    {
+      "epoch": 0.8300204850274449,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018793050269724305,
+      "loss": 2.8139,
+      "step": 16277
+    },
+    {
+      "epoch": 0.8300714784835502,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001878951197227717,
+      "loss": 2.8268,
+      "step": 16278
+    },
+    {
+      "epoch": 0.8301224719396556,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001878596886986768,
+      "loss": 2.8219,
+      "step": 16279
+    },
+    {
+      "epoch": 0.830173465395761,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018782420964681397,
+      "loss": 2.798,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8302244588518664,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018778868258906847,
+      "loss": 2.8127,
+      "step": 16281
+    },
+    {
+      "epoch": 0.8302754523079718,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0018775310754735517,
+      "loss": 2.7717,
+      "step": 16282
+    },
+    {
+      "epoch": 0.8303264457640772,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0018771748454361858,
+      "loss": 2.8124,
+      "step": 16283
+    },
+    {
+      "epoch": 0.8303774392201826,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018768181359983273,
+      "loss": 2.8048,
+      "step": 16284
+    },
+    {
+      "epoch": 0.830428432676288,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0018764609473800127,
+      "loss": 2.801,
+      "step": 16285
+    },
+    {
+      "epoch": 0.8304794261323933,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018761032798015737,
+      "loss": 2.8139,
+      "step": 16286
+    },
+    {
+      "epoch": 0.8305304195884987,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0018757451334836376,
+      "loss": 2.8382,
+      "step": 16287
+    },
+    {
+      "epoch": 0.8305814130446041,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018753865086471272,
+      "loss": 2.8292,
+      "step": 16288
+    },
+    {
+      "epoch": 0.8306324065007096,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0018750274055132602,
+      "loss": 2.8078,
+      "step": 16289
+    },
+    {
+      "epoch": 0.830683399956815,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0018746678243035492,
+      "loss": 2.8301,
+      "step": 16290
+    },
+    {
+      "epoch": 0.8307343934129203,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018743077652398027,
+      "loss": 2.7933,
+      "step": 16291
+    },
+    {
+      "epoch": 0.8307853868690257,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0018739472285441226,
+      "loss": 2.8088,
+      "step": 16292
+    },
+    {
+      "epoch": 0.8308363803251311,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0018735862144389063,
+      "loss": 2.8079,
+      "step": 16293
+    },
+    {
+      "epoch": 0.8308873737812364,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0018732247231468453,
+      "loss": 2.7902,
+      "step": 16294
+    },
+    {
+      "epoch": 0.8309383672373418,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0018728627548909255,
+      "loss": 2.8156,
+      "step": 16295
+    },
+    {
+      "epoch": 0.8309893606934473,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0018725003098944274,
+      "loss": 2.8263,
+      "step": 16296
+    },
+    {
+      "epoch": 0.8310403541495527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0018721373883809251,
+      "loss": 2.7586,
+      "step": 16297
+    },
+    {
+      "epoch": 0.8310913476056581,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0018717739905742873,
+      "loss": 2.8103,
+      "step": 16298
+    },
+    {
+      "epoch": 0.8311423410617634,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001871410116698675,
+      "loss": 2.824,
+      "step": 16299
+    },
+    {
+      "epoch": 0.8311933345178688,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018710457669785447,
+      "loss": 2.813,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8312443279739742,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018706809416386455,
+      "loss": 2.811,
+      "step": 16301
+    },
+    {
+      "epoch": 0.8312953214300796,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0018703156409040197,
+      "loss": 2.7906,
+      "step": 16302
+    },
+    {
+      "epoch": 0.831346314886185,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0018699498650000034,
+      "loss": 2.8132,
+      "step": 16303
+    },
+    {
+      "epoch": 0.8313973083422904,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0018695836141522256,
+      "loss": 2.7963,
+      "step": 16304
+    },
+    {
+      "epoch": 0.8314483017983958,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001869216888586608,
+      "loss": 2.8021,
+      "step": 16305
+    },
+    {
+      "epoch": 0.8314992952545012,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018688496885293658,
+      "loss": 2.8221,
+      "step": 16306
+    },
+    {
+      "epoch": 0.8315502887106065,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0018684820142070059,
+      "loss": 2.7955,
+      "step": 16307
+    },
+    {
+      "epoch": 0.8316012821667119,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0018681138658463285,
+      "loss": 2.791,
+      "step": 16308
+    },
+    {
+      "epoch": 0.8316522756228173,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001867745243674426,
+      "loss": 2.8119,
+      "step": 16309
+    },
+    {
+      "epoch": 0.8317032690789228,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001867376147918683,
+      "loss": 2.8065,
+      "step": 16310
+    },
+    {
+      "epoch": 0.8317542625350282,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018670065788067764,
+      "loss": 2.8152,
+      "step": 16311
+    },
+    {
+      "epoch": 0.8318052559911335,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0018666365365666753,
+      "loss": 2.807,
+      "step": 16312
+    },
+    {
+      "epoch": 0.8318562494472389,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00186626602142664,
+      "loss": 2.8,
+      "step": 16313
+    },
+    {
+      "epoch": 0.8319072429033443,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018658950336152228,
+      "loss": 2.8159,
+      "step": 16314
+    },
+    {
+      "epoch": 0.8319582363594497,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018655235733612678,
+      "loss": 2.7977,
+      "step": 16315
+    },
+    {
+      "epoch": 0.832009229815555,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00186515164089391,
+      "loss": 2.8136,
+      "step": 16316
+    },
+    {
+      "epoch": 0.8320602232716605,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018647792364425767,
+      "loss": 2.8159,
+      "step": 16317
+    },
+    {
+      "epoch": 0.8321112167277659,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0018644063602369849,
+      "loss": 2.8173,
+      "step": 16318
+    },
+    {
+      "epoch": 0.8321622101838713,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001864033012507144,
+      "loss": 2.8218,
+      "step": 16319
+    },
+    {
+      "epoch": 0.8322132036399766,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018636591934833536,
+      "loss": 2.7973,
+      "step": 16320
+    },
+    {
+      "epoch": 0.832264197096082,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0018632849033962033,
+      "loss": 2.8093,
+      "step": 16321
+    },
+    {
+      "epoch": 0.8323151905521874,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018629101424765748,
+      "loss": 2.7797,
+      "step": 16322
+    },
+    {
+      "epoch": 0.8323661840082928,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0018625349109556398,
+      "loss": 2.8309,
+      "step": 16323
+    },
+    {
+      "epoch": 0.8324171774643983,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0018621592090648587,
+      "loss": 2.8173,
+      "step": 16324
+    },
+    {
+      "epoch": 0.8324681709205036,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018617830370359842,
+      "loss": 2.8212,
+      "step": 16325
+    },
+    {
+      "epoch": 0.832519164376609,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0018614063951010581,
+      "loss": 2.8129,
+      "step": 16326
+    },
+    {
+      "epoch": 0.8325701578327144,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018610292834924116,
+      "loss": 2.8092,
+      "step": 16327
+    },
+    {
+      "epoch": 0.8326211512888197,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018606517024426666,
+      "loss": 2.8072,
+      "step": 16328
+    },
+    {
+      "epoch": 0.8326721447449251,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0018602736521847337,
+      "loss": 2.8064,
+      "step": 16329
+    },
+    {
+      "epoch": 0.8327231382010305,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018598951329518137,
+      "loss": 2.833,
+      "step": 16330
+    },
+    {
+      "epoch": 0.832774131657136,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018595161449773958,
+      "loss": 2.7881,
+      "step": 16331
+    },
+    {
+      "epoch": 0.8328251251132414,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001859136688495259,
+      "loss": 2.8611,
+      "step": 16332
+    },
+    {
+      "epoch": 0.8328761185693467,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0018587567637394714,
+      "loss": 2.796,
+      "step": 16333
+    },
+    {
+      "epoch": 0.8329271120254521,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001858376370944389,
+      "loss": 2.7998,
+      "step": 16334
+    },
+    {
+      "epoch": 0.8329781054815575,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.001857995510344658,
+      "loss": 2.8192,
+      "step": 16335
+    },
+    {
+      "epoch": 0.8330290989376629,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0018576141821752118,
+      "loss": 2.8153,
+      "step": 16336
+    },
+    {
+      "epoch": 0.8330800923937682,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001857232386671273,
+      "loss": 2.8102,
+      "step": 16337
+    },
+    {
+      "epoch": 0.8331310858498737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0018568501240683521,
+      "loss": 2.8414,
+      "step": 16338
+    },
+    {
+      "epoch": 0.8331820793059791,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0018564673946022483,
+      "loss": 2.7903,
+      "step": 16339
+    },
+    {
+      "epoch": 0.8332330727620845,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0018560841985090478,
+      "loss": 2.8209,
+      "step": 16340
+    },
+    {
+      "epoch": 0.8332840662181898,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0018557005360251254,
+      "loss": 2.8152,
+      "step": 16341
+    },
+    {
+      "epoch": 0.8333350596742952,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0018553164073871434,
+      "loss": 2.784,
+      "step": 16342
+    },
+    {
+      "epoch": 0.8333860531304006,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0018549318128320523,
+      "loss": 2.7748,
+      "step": 16343
+    },
+    {
+      "epoch": 0.833437046586506,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0018545467525970878,
+      "loss": 2.7996,
+      "step": 16344
+    },
+    {
+      "epoch": 0.8334880400426115,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0018541612269197762,
+      "loss": 2.7907,
+      "step": 16345
+    },
+    {
+      "epoch": 0.8335390334987168,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0018537752360379277,
+      "loss": 2.8106,
+      "step": 16346
+    },
+    {
+      "epoch": 0.8335900269548222,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018533887801896415,
+      "loss": 2.794,
+      "step": 16347
+    },
+    {
+      "epoch": 0.8336410204109276,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0018530018596133028,
+      "loss": 2.7781,
+      "step": 16348
+    },
+    {
+      "epoch": 0.833692013867033,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001852614474547584,
+      "loss": 2.8174,
+      "step": 16349
+    },
+    {
+      "epoch": 0.8337430073231383,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0018522266252314434,
+      "loss": 2.8046,
+      "step": 16350
+    },
+    {
+      "epoch": 0.8337940007792437,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0018518383119041258,
+      "loss": 2.7832,
+      "step": 16351
+    },
+    {
+      "epoch": 0.8338449942353492,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0018514495348051626,
+      "loss": 2.8088,
+      "step": 16352
+    },
+    {
+      "epoch": 0.8338959876914546,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018510602941743711,
+      "loss": 2.825,
+      "step": 16353
+    },
+    {
+      "epoch": 0.8339469811475599,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0018506705902518541,
+      "loss": 2.8199,
+      "step": 16354
+    },
+    {
+      "epoch": 0.8339979746036653,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0018502804232780011,
+      "loss": 2.8355,
+      "step": 16355
+    },
+    {
+      "epoch": 0.8340489680597707,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018498897934934862,
+      "loss": 2.8117,
+      "step": 16356
+    },
+    {
+      "epoch": 0.8340999615158761,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018494987011392697,
+      "loss": 2.8,
+      "step": 16357
+    },
+    {
+      "epoch": 0.8341509549719814,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001849107146456597,
+      "loss": 2.8244,
+      "step": 16358
+    },
+    {
+      "epoch": 0.8342019484280869,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001848715129686999,
+      "loss": 2.8383,
+      "step": 16359
+    },
+    {
+      "epoch": 0.8342529418841923,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0018483226510722908,
+      "loss": 2.7927,
+      "step": 16360
+    },
+    {
+      "epoch": 0.8343039353402977,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0018479297108545732,
+      "loss": 2.8204,
+      "step": 16361
+    },
+    {
+      "epoch": 0.834354928796403,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018475363092762315,
+      "loss": 2.8098,
+      "step": 16362
+    },
+    {
+      "epoch": 0.8344059222525084,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018471424465799353,
+      "loss": 2.8154,
+      "step": 16363
+    },
+    {
+      "epoch": 0.8344569157086138,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0018467481230086394,
+      "loss": 2.7908,
+      "step": 16364
+    },
+    {
+      "epoch": 0.8345079091647192,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018463533388055817,
+      "loss": 2.7907,
+      "step": 16365
+    },
+    {
+      "epoch": 0.8345589026208247,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018459580942142857,
+      "loss": 2.8102,
+      "step": 16366
+    },
+    {
+      "epoch": 0.83460989607693,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0018455623894785575,
+      "loss": 2.8043,
+      "step": 16367
+    },
+    {
+      "epoch": 0.8346608895330354,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0018451662248424882,
+      "loss": 2.817,
+      "step": 16368
+    },
+    {
+      "epoch": 0.8347118829891408,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0018447696005504518,
+      "loss": 2.8118,
+      "step": 16369
+    },
+    {
+      "epoch": 0.8347628764452462,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0018443725168471054,
+      "loss": 2.8118,
+      "step": 16370
+    },
+    {
+      "epoch": 0.8348138699013515,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0018439749739773918,
+      "loss": 2.8178,
+      "step": 16371
+    },
+    {
+      "epoch": 0.8348648633574569,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0018435769721865337,
+      "loss": 2.7934,
+      "step": 16372
+    },
+    {
+      "epoch": 0.8349158568135624,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018431785117200394,
+      "loss": 2.8019,
+      "step": 16373
+    },
+    {
+      "epoch": 0.8349668502696678,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018427795928236998,
+      "loss": 2.8001,
+      "step": 16374
+    },
+    {
+      "epoch": 0.8350178437257731,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001842380215743587,
+      "loss": 2.8438,
+      "step": 16375
+    },
+    {
+      "epoch": 0.8350688371818785,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0018419803807260577,
+      "loss": 2.8172,
+      "step": 16376
+    },
+    {
+      "epoch": 0.8351198306379839,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00184158008801775,
+      "loss": 2.8243,
+      "step": 16377
+    },
+    {
+      "epoch": 0.8351708240940893,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0018411793378655845,
+      "loss": 2.8135,
+      "step": 16378
+    },
+    {
+      "epoch": 0.8352218175501946,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0018407781305167644,
+      "loss": 2.8136,
+      "step": 16379
+    },
+    {
+      "epoch": 0.8352728110063001,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0018403764662187738,
+      "loss": 2.8053,
+      "step": 16380
+    },
+    {
+      "epoch": 0.8353238044624055,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00183997434521938,
+      "loss": 2.8141,
+      "step": 16381
+    },
+    {
+      "epoch": 0.8353747979185109,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0018395717677666312,
+      "loss": 2.8112,
+      "step": 16382
+    },
+    {
+      "epoch": 0.8354257913746163,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0018391687341088577,
+      "loss": 2.8305,
+      "step": 16383
+    },
+    {
+      "epoch": 0.8354767848307216,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0018387652444946704,
+      "loss": 2.8461,
+      "step": 16384
+    },
+    {
+      "epoch": 0.835527778286827,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0018383612991729624,
+      "loss": 2.7758,
+      "step": 16385
+    },
+    {
+      "epoch": 0.8355787717429324,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018379568983929071,
+      "loss": 2.7912,
+      "step": 16386
+    },
+    {
+      "epoch": 0.8356297651990379,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0018375520424039598,
+      "loss": 2.8202,
+      "step": 16387
+    },
+    {
+      "epoch": 0.8356807586551432,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018371467314558555,
+      "loss": 2.7919,
+      "step": 16388
+    },
+    {
+      "epoch": 0.8357317521112486,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018367409657986103,
+      "loss": 2.8226,
+      "step": 16389
+    },
+    {
+      "epoch": 0.835782745567354,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001836334745682521,
+      "loss": 2.777,
+      "step": 16390
+    },
+    {
+      "epoch": 0.8358337390234594,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.001835928071358165,
+      "loss": 2.7936,
+      "step": 16391
+    },
+    {
+      "epoch": 0.8358847324795647,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.001835520943076399,
+      "loss": 2.8196,
+      "step": 16392
+    },
+    {
+      "epoch": 0.8359357259356701,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0018351133610883601,
+      "loss": 2.7844,
+      "step": 16393
+    },
+    {
+      "epoch": 0.8359867193917756,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018347053256454657,
+      "loss": 2.8222,
+      "step": 16394
+    },
+    {
+      "epoch": 0.836037712847881,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0018342968369994123,
+      "loss": 2.7828,
+      "step": 16395
+    },
+    {
+      "epoch": 0.8360887063039864,
+      "grad_norm": 5.71875,
+      "learning_rate": 0.001833887895402176,
+      "loss": 2.8345,
+      "step": 16396
+    },
+    {
+      "epoch": 0.8361396997600917,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0018334785011060134,
+      "loss": 2.8155,
+      "step": 16397
+    },
+    {
+      "epoch": 0.8361906932161971,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0018330686543634586,
+      "loss": 2.808,
+      "step": 16398
+    },
+    {
+      "epoch": 0.8362416866723025,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0018326583554273256,
+      "loss": 2.8306,
+      "step": 16399
+    },
+    {
+      "epoch": 0.8362926801284079,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018322476045507083,
+      "loss": 2.7871,
+      "step": 16400
+    },
+    {
+      "epoch": 0.8363436735845133,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018318364019869777,
+      "loss": 2.7987,
+      "step": 16401
+    },
+    {
+      "epoch": 0.8363946670406187,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0018314247479897846,
+      "loss": 2.8281,
+      "step": 16402
+    },
+    {
+      "epoch": 0.8364456604967241,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018310126428130575,
+      "loss": 2.8203,
+      "step": 16403
+    },
+    {
+      "epoch": 0.8364966539528295,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001830600086711004,
+      "loss": 2.8026,
+      "step": 16404
+    },
+    {
+      "epoch": 0.8365476474089348,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018301870799381091,
+      "loss": 2.7937,
+      "step": 16405
+    },
+    {
+      "epoch": 0.8365986408650402,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0018297736227491361,
+      "loss": 2.8118,
+      "step": 16406
+    },
+    {
+      "epoch": 0.8366496343211456,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0018293597153991265,
+      "loss": 2.812,
+      "step": 16407
+    },
+    {
+      "epoch": 0.8367006277772511,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0018289453581433988,
+      "loss": 2.8352,
+      "step": 16408
+    },
+    {
+      "epoch": 0.8367516212333564,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0018285305512375498,
+      "loss": 2.8119,
+      "step": 16409
+    },
+    {
+      "epoch": 0.8368026146894618,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0018281152949374527,
+      "loss": 2.7895,
+      "step": 16410
+    },
+    {
+      "epoch": 0.8368536081455672,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001827699589499259,
+      "loss": 2.8181,
+      "step": 16411
+    },
+    {
+      "epoch": 0.8369046016016726,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001827283435179396,
+      "loss": 2.7854,
+      "step": 16412
+    },
+    {
+      "epoch": 0.836955595057778,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0018268668322345694,
+      "loss": 2.8091,
+      "step": 16413
+    },
+    {
+      "epoch": 0.8370065885138833,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0018264497809217602,
+      "loss": 2.7851,
+      "step": 16414
+    },
+    {
+      "epoch": 0.8370575819699888,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018260322814982269,
+      "loss": 2.8073,
+      "step": 16415
+    },
+    {
+      "epoch": 0.8371085754260942,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0018256143342215043,
+      "loss": 2.789,
+      "step": 16416
+    },
+    {
+      "epoch": 0.8371595688821996,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0018251959393494031,
+      "loss": 2.7847,
+      "step": 16417
+    },
+    {
+      "epoch": 0.8372105623383049,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018247770971400097,
+      "loss": 2.8019,
+      "step": 16418
+    },
+    {
+      "epoch": 0.8372615557944103,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001824357807851688,
+      "loss": 2.8092,
+      "step": 16419
+    },
+    {
+      "epoch": 0.8373125492505157,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001823938071743076,
+      "loss": 2.8153,
+      "step": 16420
+    },
+    {
+      "epoch": 0.8373635427066211,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0018235178890730886,
+      "loss": 2.7941,
+      "step": 16421
+    },
+    {
+      "epoch": 0.8374145361627265,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001823097260100915,
+      "loss": 2.7601,
+      "step": 16422
+    },
+    {
+      "epoch": 0.8374655296188319,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018226761850860206,
+      "loss": 2.833,
+      "step": 16423
+    },
+    {
+      "epoch": 0.8375165230749373,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018222546642881453,
+      "loss": 2.8039,
+      "step": 16424
+    },
+    {
+      "epoch": 0.8375675165310427,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001821832697967305,
+      "loss": 2.7986,
+      "step": 16425
+    },
+    {
+      "epoch": 0.837618509987148,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.001821410286383789,
+      "loss": 2.8212,
+      "step": 16426
+    },
+    {
+      "epoch": 0.8376695034432534,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018209874297981622,
+      "loss": 2.8117,
+      "step": 16427
+    },
+    {
+      "epoch": 0.8377204968993588,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0018205641284712642,
+      "loss": 2.8236,
+      "step": 16428
+    },
+    {
+      "epoch": 0.8377714903554643,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001820140382664208,
+      "loss": 2.7863,
+      "step": 16429
+    },
+    {
+      "epoch": 0.8378224838115697,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0018197161926383817,
+      "loss": 2.825,
+      "step": 16430
+    },
+    {
+      "epoch": 0.837873477267675,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0018192915586554466,
+      "loss": 2.8079,
+      "step": 16431
+    },
+    {
+      "epoch": 0.8379244707237804,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018188664809773384,
+      "loss": 2.8039,
+      "step": 16432
+    },
+    {
+      "epoch": 0.8379754641798858,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0018184409598662667,
+      "loss": 2.795,
+      "step": 16433
+    },
+    {
+      "epoch": 0.8380264576359912,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0018180149955847138,
+      "loss": 2.7991,
+      "step": 16434
+    },
+    {
+      "epoch": 0.8380774510920965,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0018175885883954363,
+      "loss": 2.8189,
+      "step": 16435
+    },
+    {
+      "epoch": 0.838128444548202,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0018171617385614636,
+      "loss": 2.7815,
+      "step": 16436
+    },
+    {
+      "epoch": 0.8381794380043074,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0018167344463460976,
+      "loss": 2.8292,
+      "step": 16437
+    },
+    {
+      "epoch": 0.8382304314604128,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0018163067120129143,
+      "loss": 2.7995,
+      "step": 16438
+    },
+    {
+      "epoch": 0.8382814249165181,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018158785358257615,
+      "loss": 2.8083,
+      "step": 16439
+    },
+    {
+      "epoch": 0.8383324183726235,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0018154499180487592,
+      "loss": 2.8233,
+      "step": 16440
+    },
+    {
+      "epoch": 0.8383834118287289,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.001815020858946301,
+      "loss": 2.791,
+      "step": 16441
+    },
+    {
+      "epoch": 0.8384344052848343,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0018145913587830518,
+      "loss": 2.805,
+      "step": 16442
+    },
+    {
+      "epoch": 0.8384853987409397,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018141614178239492,
+      "loss": 2.8041,
+      "step": 16443
+    },
+    {
+      "epoch": 0.8385363921970451,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0018137310363342022,
+      "loss": 2.8176,
+      "step": 16444
+    },
+    {
+      "epoch": 0.8385873856531505,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001813300214579292,
+      "loss": 2.7892,
+      "step": 16445
+    },
+    {
+      "epoch": 0.8386383791092559,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0018128689528249704,
+      "loss": 2.8285,
+      "step": 16446
+    },
+    {
+      "epoch": 0.8386893725653612,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001812437251337262,
+      "loss": 2.8015,
+      "step": 16447
+    },
+    {
+      "epoch": 0.8387403660214666,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0018120051103824618,
+      "loss": 2.8017,
+      "step": 16448
+    },
+    {
+      "epoch": 0.838791359477572,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018115725302271354,
+      "loss": 2.8031,
+      "step": 16449
+    },
+    {
+      "epoch": 0.8388423529336775,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0018111395111381214,
+      "loss": 2.8151,
+      "step": 16450
+    },
+    {
+      "epoch": 0.8388933463897829,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0018107060533825265,
+      "loss": 2.8312,
+      "step": 16451
+    },
+    {
+      "epoch": 0.8389443398458882,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0018102721572277297,
+      "loss": 2.8348,
+      "step": 16452
+    },
+    {
+      "epoch": 0.8389953333019936,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0018098378229413799,
+      "loss": 2.7859,
+      "step": 16453
+    },
+    {
+      "epoch": 0.839046326758099,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001809403050791396,
+      "loss": 2.8174,
+      "step": 16454
+    },
+    {
+      "epoch": 0.8390973202142044,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0018089678410459678,
+      "loss": 2.8112,
+      "step": 16455
+    },
+    {
+      "epoch": 0.8391483136703097,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0018085321939735543,
+      "loss": 2.8005,
+      "step": 16456
+    },
+    {
+      "epoch": 0.8391993071264152,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018080961098428845,
+      "loss": 2.8163,
+      "step": 16457
+    },
+    {
+      "epoch": 0.8392503005825206,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001807659588922957,
+      "loss": 2.7888,
+      "step": 16458
+    },
+    {
+      "epoch": 0.839301294038626,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00180722263148304,
+      "loss": 2.7456,
+      "step": 16459
+    },
+    {
+      "epoch": 0.8393522874947313,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0018067852377926703,
+      "loss": 2.805,
+      "step": 16460
+    },
+    {
+      "epoch": 0.8394032809508367,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0018063474081216554,
+      "loss": 2.7833,
+      "step": 16461
+    },
+    {
+      "epoch": 0.8394542744069421,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018059091427400692,
+      "loss": 2.8109,
+      "step": 16462
+    },
+    {
+      "epoch": 0.8395052678630475,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018054704419182565,
+      "loss": 2.7896,
+      "step": 16463
+    },
+    {
+      "epoch": 0.839556261319153,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018050313059268304,
+      "loss": 2.8096,
+      "step": 16464
+    },
+    {
+      "epoch": 0.8396072547752583,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018045917350366716,
+      "loss": 2.818,
+      "step": 16465
+    },
+    {
+      "epoch": 0.8396582482313637,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0018041517295189291,
+      "loss": 2.8046,
+      "step": 16466
+    },
+    {
+      "epoch": 0.8397092416874691,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018037112896450215,
+      "loss": 2.7943,
+      "step": 16467
+    },
+    {
+      "epoch": 0.8397602351435745,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0018032704156866334,
+      "loss": 2.7963,
+      "step": 16468
+    },
+    {
+      "epoch": 0.8398112285996798,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018028291079157182,
+      "loss": 2.7724,
+      "step": 16469
+    },
+    {
+      "epoch": 0.8398622220557852,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0018023873666044967,
+      "loss": 2.8191,
+      "step": 16470
+    },
+    {
+      "epoch": 0.8399132155118907,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0018019451920254575,
+      "loss": 2.8265,
+      "step": 16471
+    },
+    {
+      "epoch": 0.8399642089679961,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0018015025844513555,
+      "loss": 2.8052,
+      "step": 16472
+    },
+    {
+      "epoch": 0.8400152024241014,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018010595441552133,
+      "loss": 2.7724,
+      "step": 16473
+    },
+    {
+      "epoch": 0.8400661958802068,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0018006160714103213,
+      "loss": 2.8012,
+      "step": 16474
+    },
+    {
+      "epoch": 0.8401171893363122,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0018001721664902346,
+      "loss": 2.7727,
+      "step": 16475
+    },
+    {
+      "epoch": 0.8401681827924176,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001799727829668777,
+      "loss": 2.7845,
+      "step": 16476
+    },
+    {
+      "epoch": 0.8402191762485229,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0017992830612200371,
+      "loss": 2.772,
+      "step": 16477
+    },
+    {
+      "epoch": 0.8402701697046284,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017988378614183706,
+      "loss": 2.792,
+      "step": 16478
+    },
+    {
+      "epoch": 0.8403211631607338,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0017983922305383998,
+      "loss": 2.8054,
+      "step": 16479
+    },
+    {
+      "epoch": 0.8403721566168392,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0017979461688550108,
+      "loss": 2.8081,
+      "step": 16480
+    },
+    {
+      "epoch": 0.8404231500729445,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017974996766433585,
+      "loss": 2.8285,
+      "step": 16481
+    },
+    {
+      "epoch": 0.8404741435290499,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0017970527541788604,
+      "loss": 2.8072,
+      "step": 16482
+    },
+    {
+      "epoch": 0.8405251369851553,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017966054017372016,
+      "loss": 2.8162,
+      "step": 16483
+    },
+    {
+      "epoch": 0.8405761304412607,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.001796157619594331,
+      "loss": 2.8012,
+      "step": 16484
+    },
+    {
+      "epoch": 0.8406271238973662,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0017957094080264632,
+      "loss": 2.817,
+      "step": 16485
+    },
+    {
+      "epoch": 0.8406781173534715,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.001795260767310078,
+      "loss": 2.7851,
+      "step": 16486
+    },
+    {
+      "epoch": 0.8407291108095769,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0017948116977219194,
+      "loss": 2.8108,
+      "step": 16487
+    },
+    {
+      "epoch": 0.8407801042656823,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0017943621995389957,
+      "loss": 2.8021,
+      "step": 16488
+    },
+    {
+      "epoch": 0.8408310977217877,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017939122730385802,
+      "loss": 2.8063,
+      "step": 16489
+    },
+    {
+      "epoch": 0.840882091177893,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0017934619184982103,
+      "loss": 2.8219,
+      "step": 16490
+    },
+    {
+      "epoch": 0.8409330846339984,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0017930111361956873,
+      "loss": 2.7925,
+      "step": 16491
+    },
+    {
+      "epoch": 0.8409840780901039,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0017925599264090762,
+      "loss": 2.784,
+      "step": 16492
+    },
+    {
+      "epoch": 0.8410350715462093,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001792108289416706,
+      "loss": 2.7926,
+      "step": 16493
+    },
+    {
+      "epoch": 0.8410860650023146,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0017916562254971689,
+      "loss": 2.7937,
+      "step": 16494
+    },
+    {
+      "epoch": 0.84113705845842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.001791203734929321,
+      "loss": 2.7908,
+      "step": 16495
+    },
+    {
+      "epoch": 0.8411880519145254,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001790750817992281,
+      "loss": 2.8064,
+      "step": 16496
+    },
+    {
+      "epoch": 0.8412390453706308,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0017902974749654306,
+      "loss": 2.8037,
+      "step": 16497
+    },
+    {
+      "epoch": 0.8412900388267361,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001789843706128415,
+      "loss": 2.7976,
+      "step": 16498
+    },
+    {
+      "epoch": 0.8413410322828416,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017893895117611412,
+      "loss": 2.7792,
+      "step": 16499
+    },
+    {
+      "epoch": 0.841392025738947,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0017889348921437793,
+      "loss": 2.7956,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8414430191950524,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0017884798475567617,
+      "loss": 2.7967,
+      "step": 16501
+    },
+    {
+      "epoch": 0.8414940126511578,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0017880243782807826,
+      "loss": 2.8033,
+      "step": 16502
+    },
+    {
+      "epoch": 0.8415450061072631,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.001787568484596798,
+      "loss": 2.7897,
+      "step": 16503
+    },
+    {
+      "epoch": 0.8415959995633685,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017871121667860268,
+      "loss": 2.7929,
+      "step": 16504
+    },
+    {
+      "epoch": 0.8416469930194739,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0017866554251299478,
+      "loss": 2.8017,
+      "step": 16505
+    },
+    {
+      "epoch": 0.8416979864755794,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001786198259910303,
+      "loss": 2.7678,
+      "step": 16506
+    },
+    {
+      "epoch": 0.8417489799316847,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0017857406714090946,
+      "loss": 2.8104,
+      "step": 16507
+    },
+    {
+      "epoch": 0.8417999733877901,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017852826599085863,
+      "loss": 2.8051,
+      "step": 16508
+    },
+    {
+      "epoch": 0.8418509668438955,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0017848242256913022,
+      "loss": 2.8033,
+      "step": 16509
+    },
+    {
+      "epoch": 0.8419019603000009,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.001784365369040028,
+      "loss": 2.8162,
+      "step": 16510
+    },
+    {
+      "epoch": 0.8419529537561062,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017839060902378092,
+      "loss": 2.7909,
+      "step": 16511
+    },
+    {
+      "epoch": 0.8420039472122116,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.001783446389567952,
+      "loss": 2.8075,
+      "step": 16512
+    },
+    {
+      "epoch": 0.8420549406683171,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0017829862673140239,
+      "loss": 2.8026,
+      "step": 16513
+    },
+    {
+      "epoch": 0.8421059341244225,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017825257237598499,
+      "loss": 2.7946,
+      "step": 16514
+    },
+    {
+      "epoch": 0.8421569275805278,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0017820647591895171,
+      "loss": 2.8032,
+      "step": 16515
+    },
+    {
+      "epoch": 0.8422079210366332,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017816033738873717,
+      "loss": 2.8015,
+      "step": 16516
+    },
+    {
+      "epoch": 0.8422589144927386,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017811415681380195,
+      "loss": 2.7839,
+      "step": 16517
+    },
+    {
+      "epoch": 0.842309907948844,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017806793422263248,
+      "loss": 2.8045,
+      "step": 16518
+    },
+    {
+      "epoch": 0.8423609014049493,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0017802166964374126,
+      "loss": 2.8052,
+      "step": 16519
+    },
+    {
+      "epoch": 0.8424118948610548,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0017797536310566653,
+      "loss": 2.8316,
+      "step": 16520
+    },
+    {
+      "epoch": 0.8424628883171602,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0017792901463697254,
+      "loss": 2.8515,
+      "step": 16521
+    },
+    {
+      "epoch": 0.8425138817732656,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0017788262426624935,
+      "loss": 2.8029,
+      "step": 16522
+    },
+    {
+      "epoch": 0.842564875229371,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0017783619202211285,
+      "loss": 2.8126,
+      "step": 16523
+    },
+    {
+      "epoch": 0.8426158686854763,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001777897179332048,
+      "loss": 2.7941,
+      "step": 16524
+    },
+    {
+      "epoch": 0.8426668621415817,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0017774320202819274,
+      "loss": 2.823,
+      "step": 16525
+    },
+    {
+      "epoch": 0.8427178555976871,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0017769664433577002,
+      "loss": 2.772,
+      "step": 16526
+    },
+    {
+      "epoch": 0.8427688490537926,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0017765004488465577,
+      "loss": 2.8148,
+      "step": 16527
+    },
+    {
+      "epoch": 0.8428198425098979,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0017760340370359489,
+      "loss": 2.7908,
+      "step": 16528
+    },
+    {
+      "epoch": 0.8428708359660033,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0017755672082135797,
+      "loss": 2.7807,
+      "step": 16529
+    },
+    {
+      "epoch": 0.8429218294221087,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0017750999626674135,
+      "loss": 2.7927,
+      "step": 16530
+    },
+    {
+      "epoch": 0.8429728228782141,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0017746323006856712,
+      "loss": 2.7951,
+      "step": 16531
+    },
+    {
+      "epoch": 0.8430238163343194,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017741642225568303,
+      "loss": 2.8008,
+      "step": 16532
+    },
+    {
+      "epoch": 0.8430748097904248,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0017736957285696242,
+      "loss": 2.8285,
+      "step": 16533
+    },
+    {
+      "epoch": 0.8431258032465303,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017732268190130444,
+      "loss": 2.8162,
+      "step": 16534
+    },
+    {
+      "epoch": 0.8431767967026357,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017727574941763372,
+      "loss": 2.7891,
+      "step": 16535
+    },
+    {
+      "epoch": 0.843227790158741,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001772287754349006,
+      "loss": 2.8141,
+      "step": 16536
+    },
+    {
+      "epoch": 0.8432787836148464,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00177181759982081,
+      "loss": 2.7939,
+      "step": 16537
+    },
+    {
+      "epoch": 0.8433297770709518,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.001771347030881764,
+      "loss": 2.7961,
+      "step": 16538
+    },
+    {
+      "epoch": 0.8433807705270572,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0017708760478221386,
+      "loss": 2.8313,
+      "step": 16539
+    },
+    {
+      "epoch": 0.8434317639831626,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017704046509324598,
+      "loss": 2.7641,
+      "step": 16540
+    },
+    {
+      "epoch": 0.843482757439268,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0017699328405035091,
+      "loss": 2.7882,
+      "step": 16541
+    },
+    {
+      "epoch": 0.8435337508953734,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017694606168263225,
+      "loss": 2.7902,
+      "step": 16542
+    },
+    {
+      "epoch": 0.8435847443514788,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0017689879801921916,
+      "loss": 2.7872,
+      "step": 16543
+    },
+    {
+      "epoch": 0.8436357378075842,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0017685149308926625,
+      "loss": 2.8066,
+      "step": 16544
+    },
+    {
+      "epoch": 0.8436867312636895,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0017680414692195353,
+      "loss": 2.7965,
+      "step": 16545
+    },
+    {
+      "epoch": 0.8437377247197949,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0017675675954648654,
+      "loss": 2.7993,
+      "step": 16546
+    },
+    {
+      "epoch": 0.8437887181759003,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0017670933099209615,
+      "loss": 2.8079,
+      "step": 16547
+    },
+    {
+      "epoch": 0.8438397116320058,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017666186128803877,
+      "loss": 2.779,
+      "step": 16548
+    },
+    {
+      "epoch": 0.8438907050881111,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0017661435046359595,
+      "loss": 2.7812,
+      "step": 16549
+    },
+    {
+      "epoch": 0.8439416985442165,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001765667985480749,
+      "loss": 2.7976,
+      "step": 16550
+    },
+    {
+      "epoch": 0.8439926920003219,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017651920557080793,
+      "loss": 2.8207,
+      "step": 16551
+    },
+    {
+      "epoch": 0.8440436854564273,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0017647157156115284,
+      "loss": 2.8172,
+      "step": 16552
+    },
+    {
+      "epoch": 0.8440946789125326,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.001764238965484926,
+      "loss": 2.8013,
+      "step": 16553
+    },
+    {
+      "epoch": 0.844145672368638,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0017637618056223566,
+      "loss": 2.8399,
+      "step": 16554
+    },
+    {
+      "epoch": 0.8441966658247435,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0017632842363181557,
+      "loss": 2.8109,
+      "step": 16555
+    },
+    {
+      "epoch": 0.8442476592808489,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0017628062578669123,
+      "loss": 2.7962,
+      "step": 16556
+    },
+    {
+      "epoch": 0.8442986527369543,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017623278705634676,
+      "loss": 2.8139,
+      "step": 16557
+    },
+    {
+      "epoch": 0.8443496461930596,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017618490747029147,
+      "loss": 2.7702,
+      "step": 16558
+    },
+    {
+      "epoch": 0.844400639649165,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017613698705805988,
+      "loss": 2.7963,
+      "step": 16559
+    },
+    {
+      "epoch": 0.8444516331052704,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001760890258492117,
+      "loss": 2.7986,
+      "step": 16560
+    },
+    {
+      "epoch": 0.8445026265613758,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017604102387333187,
+      "loss": 2.7981,
+      "step": 16561
+    },
+    {
+      "epoch": 0.8445536200174812,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017599298116003042,
+      "loss": 2.8096,
+      "step": 16562
+    },
+    {
+      "epoch": 0.8446046134735866,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017594489773894244,
+      "loss": 2.8265,
+      "step": 16563
+    },
+    {
+      "epoch": 0.844655606929692,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017589677363972827,
+      "loss": 2.7843,
+      "step": 16564
+    },
+    {
+      "epoch": 0.8447066003857974,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0017584860889207324,
+      "loss": 2.8043,
+      "step": 16565
+    },
+    {
+      "epoch": 0.8447575938419027,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0017580040352568778,
+      "loss": 2.8176,
+      "step": 16566
+    },
+    {
+      "epoch": 0.8448085872980081,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0017575215757030737,
+      "loss": 2.7871,
+      "step": 16567
+    },
+    {
+      "epoch": 0.8448595807541135,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001757038710556926,
+      "loss": 2.7874,
+      "step": 16568
+    },
+    {
+      "epoch": 0.844910574210219,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017565554401162897,
+      "loss": 2.8093,
+      "step": 16569
+    },
+    {
+      "epoch": 0.8449615676663244,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0017560717646792703,
+      "loss": 2.809,
+      "step": 16570
+    },
+    {
+      "epoch": 0.8450125611224297,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017555876845442232,
+      "loss": 2.7709,
+      "step": 16571
+    },
+    {
+      "epoch": 0.8450635545785351,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0017551032000097536,
+      "loss": 2.8098,
+      "step": 16572
+    },
+    {
+      "epoch": 0.8451145480346405,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0017546183113747156,
+      "loss": 2.7817,
+      "step": 16573
+    },
+    {
+      "epoch": 0.8451655414907459,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0017541330189382125,
+      "loss": 2.8043,
+      "step": 16574
+    },
+    {
+      "epoch": 0.8452165349468512,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0017536473229995979,
+      "loss": 2.7824,
+      "step": 16575
+    },
+    {
+      "epoch": 0.8452675284029567,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017531612238584728,
+      "loss": 2.7909,
+      "step": 16576
+    },
+    {
+      "epoch": 0.8453185218590621,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017526747218146882,
+      "loss": 2.8025,
+      "step": 16577
+    },
+    {
+      "epoch": 0.8453695153151675,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001752187817168342,
+      "loss": 2.7659,
+      "step": 16578
+    },
+    {
+      "epoch": 0.8454205087712728,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0017517005102197826,
+      "loss": 2.8081,
+      "step": 16579
+    },
+    {
+      "epoch": 0.8454715022273782,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0017512128012696043,
+      "loss": 2.786,
+      "step": 16580
+    },
+    {
+      "epoch": 0.8455224956834836,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0017507246906186516,
+      "loss": 2.7772,
+      "step": 16581
+    },
+    {
+      "epoch": 0.845573489139589,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017502361785680148,
+      "loss": 2.7936,
+      "step": 16582
+    },
+    {
+      "epoch": 0.8456244825956944,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0017497472654190332,
+      "loss": 2.7817,
+      "step": 16583
+    },
+    {
+      "epoch": 0.8456754760517998,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0017492579514732927,
+      "loss": 2.8041,
+      "step": 16584
+    },
+    {
+      "epoch": 0.8457264695079052,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001748768237032627,
+      "loss": 2.7841,
+      "step": 16585
+    },
+    {
+      "epoch": 0.8457774629640106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0017482781223991163,
+      "loss": 2.8032,
+      "step": 16586
+    },
+    {
+      "epoch": 0.845828456420116,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.001747787607875088,
+      "loss": 2.7772,
+      "step": 16587
+    },
+    {
+      "epoch": 0.8458794498762213,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0017472966937631161,
+      "loss": 2.8114,
+      "step": 16588
+    },
+    {
+      "epoch": 0.8459304433323267,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017468053803660212,
+      "loss": 2.7946,
+      "step": 16589
+    },
+    {
+      "epoch": 0.8459814367884322,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0017463136679868698,
+      "loss": 2.7734,
+      "step": 16590
+    },
+    {
+      "epoch": 0.8460324302445376,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0017458215569289752,
+      "loss": 2.7984,
+      "step": 16591
+    },
+    {
+      "epoch": 0.8460834237006429,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0017453290474958962,
+      "loss": 2.7686,
+      "step": 16592
+    },
+    {
+      "epoch": 0.8461344171567483,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001744836139991437,
+      "loss": 2.7951,
+      "step": 16593
+    },
+    {
+      "epoch": 0.8461854106128537,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017443428347196479,
+      "loss": 2.8035,
+      "step": 16594
+    },
+    {
+      "epoch": 0.8462364040689591,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0017438491319848246,
+      "loss": 2.7893,
+      "step": 16595
+    },
+    {
+      "epoch": 0.8462873975250644,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0017433550320915077,
+      "loss": 2.7738,
+      "step": 16596
+    },
+    {
+      "epoch": 0.8463383909811699,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0017428605353444822,
+      "loss": 2.8083,
+      "step": 16597
+    },
+    {
+      "epoch": 0.8463893844372753,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017423656420487792,
+      "loss": 2.7704,
+      "step": 16598
+    },
+    {
+      "epoch": 0.8464403778933807,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017418703525096738,
+      "loss": 2.7962,
+      "step": 16599
+    },
+    {
+      "epoch": 0.846491371349486,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001741374667032685,
+      "loss": 2.7995,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8465423648055914,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0017408785859235768,
+      "loss": 2.8177,
+      "step": 16601
+    },
+    {
+      "epoch": 0.8465933582616968,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001740382109488357,
+      "loss": 2.7766,
+      "step": 16602
+    },
+    {
+      "epoch": 0.8466443517178022,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017398852380332767,
+      "loss": 2.7851,
+      "step": 16603
+    },
+    {
+      "epoch": 0.8466953451739077,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0017393879718648323,
+      "loss": 2.7816,
+      "step": 16604
+    },
+    {
+      "epoch": 0.846746338630013,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017388903112897612,
+      "loss": 2.8237,
+      "step": 16605
+    },
+    {
+      "epoch": 0.8467973320861184,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017383922566150465,
+      "loss": 2.7846,
+      "step": 16606
+    },
+    {
+      "epoch": 0.8468483255422238,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.001737893808147913,
+      "loss": 2.7922,
+      "step": 16607
+    },
+    {
+      "epoch": 0.8468993189983292,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0017373949661958284,
+      "loss": 2.7909,
+      "step": 16608
+    },
+    {
+      "epoch": 0.8469503124544345,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0017368957310665038,
+      "loss": 2.8178,
+      "step": 16609
+    },
+    {
+      "epoch": 0.8470013059105399,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0017363961030678928,
+      "loss": 2.8123,
+      "step": 16610
+    },
+    {
+      "epoch": 0.8470522993666454,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0017358960825081906,
+      "loss": 2.7783,
+      "step": 16611
+    },
+    {
+      "epoch": 0.8471032928227508,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001735395669695835,
+      "loss": 2.7828,
+      "step": 16612
+    },
+    {
+      "epoch": 0.8471542862788561,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017348948649395061,
+      "loss": 2.8126,
+      "step": 16613
+    },
+    {
+      "epoch": 0.8472052797349615,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0017343936685481252,
+      "loss": 2.8013,
+      "step": 16614
+    },
+    {
+      "epoch": 0.8472562731910669,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0017338920808308558,
+      "loss": 2.7809,
+      "step": 16615
+    },
+    {
+      "epoch": 0.8473072666471723,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0017333901020971016,
+      "loss": 2.8158,
+      "step": 16616
+    },
+    {
+      "epoch": 0.8473582601032776,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0017328877326565088,
+      "loss": 2.7969,
+      "step": 16617
+    },
+    {
+      "epoch": 0.8474092535593831,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017323849728189645,
+      "loss": 2.8093,
+      "step": 16618
+    },
+    {
+      "epoch": 0.8474602470154885,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017318818228945953,
+      "loss": 2.7713,
+      "step": 16619
+    },
+    {
+      "epoch": 0.8475112404715939,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0017313782831937703,
+      "loss": 2.7844,
+      "step": 16620
+    },
+    {
+      "epoch": 0.8475622339276992,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0017308743540270974,
+      "loss": 2.7981,
+      "step": 16621
+    },
+    {
+      "epoch": 0.8476132273838046,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0017303700357054258,
+      "loss": 2.7843,
+      "step": 16622
+    },
+    {
+      "epoch": 0.84766422083991,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0017298653285398443,
+      "loss": 2.7837,
+      "step": 16623
+    },
+    {
+      "epoch": 0.8477152142960154,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0017293602328416816,
+      "loss": 2.8013,
+      "step": 16624
+    },
+    {
+      "epoch": 0.8477662077521209,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001728854748922506,
+      "loss": 2.7877,
+      "step": 16625
+    },
+    {
+      "epoch": 0.8478172012082262,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017283488770941256,
+      "loss": 2.7956,
+      "step": 16626
+    },
+    {
+      "epoch": 0.8478681946643316,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0017278426176685873,
+      "loss": 2.7837,
+      "step": 16627
+    },
+    {
+      "epoch": 0.847919188120437,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017273359709581776,
+      "loss": 2.7734,
+      "step": 16628
+    },
+    {
+      "epoch": 0.8479701815765424,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017268289372754215,
+      "loss": 2.8178,
+      "step": 16629
+    },
+    {
+      "epoch": 0.8480211750326477,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0017263215169330828,
+      "loss": 2.7963,
+      "step": 16630
+    },
+    {
+      "epoch": 0.8480721684887531,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0017258137102441641,
+      "loss": 2.7986,
+      "step": 16631
+    },
+    {
+      "epoch": 0.8481231619448586,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0017253055175219059,
+      "loss": 2.8307,
+      "step": 16632
+    },
+    {
+      "epoch": 0.848174155400964,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0017247969390797873,
+      "loss": 2.8,
+      "step": 16633
+    },
+    {
+      "epoch": 0.8482251488570693,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017242879752315244,
+      "loss": 2.7927,
+      "step": 16634
+    },
+    {
+      "epoch": 0.8482761423131747,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0017237786262910724,
+      "loss": 2.7891,
+      "step": 16635
+    },
+    {
+      "epoch": 0.8483271357692801,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0017232688925726228,
+      "loss": 2.7913,
+      "step": 16636
+    },
+    {
+      "epoch": 0.8483781292253855,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0017227587743906053,
+      "loss": 2.791,
+      "step": 16637
+    },
+    {
+      "epoch": 0.8484291226814908,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0017222482720596858,
+      "loss": 2.8131,
+      "step": 16638
+    },
+    {
+      "epoch": 0.8484801161375963,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017217373858947685,
+      "loss": 2.7596,
+      "step": 16639
+    },
+    {
+      "epoch": 0.8485311095937017,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0017212261162109935,
+      "loss": 2.8026,
+      "step": 16640
+    },
+    {
+      "epoch": 0.8485821030498071,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0017207144633237376,
+      "loss": 2.795,
+      "step": 16641
+    },
+    {
+      "epoch": 0.8486330965059125,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0017202024275486136,
+      "loss": 2.7978,
+      "step": 16642
+    },
+    {
+      "epoch": 0.8486840899620178,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0017196900092014715,
+      "loss": 2.8018,
+      "step": 16643
+    },
+    {
+      "epoch": 0.8487350834181232,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0017191772085983963,
+      "loss": 2.8191,
+      "step": 16644
+    },
+    {
+      "epoch": 0.8487860768742286,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0017186640260557094,
+      "loss": 2.791,
+      "step": 16645
+    },
+    {
+      "epoch": 0.8488370703303341,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0017181504618899676,
+      "loss": 2.8111,
+      "step": 16646
+    },
+    {
+      "epoch": 0.8488880637864394,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001717636516417963,
+      "loss": 2.7653,
+      "step": 16647
+    },
+    {
+      "epoch": 0.8489390572425448,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0017171221899567222,
+      "loss": 2.7873,
+      "step": 16648
+    },
+    {
+      "epoch": 0.8489900506986502,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0017166074828235092,
+      "loss": 2.8053,
+      "step": 16649
+    },
+    {
+      "epoch": 0.8490410441547556,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0017160923953358199,
+      "loss": 2.8003,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8490920376108609,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0017155769278113867,
+      "loss": 2.7819,
+      "step": 16651
+    },
+    {
+      "epoch": 0.8491430310669663,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017150610805681758,
+      "loss": 2.7633,
+      "step": 16652
+    },
+    {
+      "epoch": 0.8491940245230718,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0017145448539243877,
+      "loss": 2.7901,
+      "step": 16653
+    },
+    {
+      "epoch": 0.8492450179791772,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0017140282481984568,
+      "loss": 2.8054,
+      "step": 16654
+    },
+    {
+      "epoch": 0.8492960114352825,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001713511263709052,
+      "loss": 2.7923,
+      "step": 16655
+    },
+    {
+      "epoch": 0.8493470048913879,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017129939007750742,
+      "loss": 2.7897,
+      "step": 16656
+    },
+    {
+      "epoch": 0.8493979983474933,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0017124761597156599,
+      "loss": 2.8006,
+      "step": 16657
+    },
+    {
+      "epoch": 0.8494489918035987,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0017119580408501774,
+      "loss": 2.7797,
+      "step": 16658
+    },
+    {
+      "epoch": 0.849499985259704,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0017114395444982286,
+      "loss": 2.802,
+      "step": 16659
+    },
+    {
+      "epoch": 0.8495509787158095,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017109206709796476,
+      "loss": 2.7909,
+      "step": 16660
+    },
+    {
+      "epoch": 0.8496019721719149,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0017104014206145023,
+      "loss": 2.8115,
+      "step": 16661
+    },
+    {
+      "epoch": 0.8496529656280203,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017098817937230921,
+      "loss": 2.79,
+      "step": 16662
+    },
+    {
+      "epoch": 0.8497039590841257,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001709361790625949,
+      "loss": 2.8146,
+      "step": 16663
+    },
+    {
+      "epoch": 0.849754952540231,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001708841411643837,
+      "loss": 2.7659,
+      "step": 16664
+    },
+    {
+      "epoch": 0.8498059459963364,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017083206570977522,
+      "loss": 2.7893,
+      "step": 16665
+    },
+    {
+      "epoch": 0.8498569394524418,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0017077995273089219,
+      "loss": 2.7717,
+      "step": 16666
+    },
+    {
+      "epoch": 0.8499079329085473,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0017072780225988056,
+      "loss": 2.7661,
+      "step": 16667
+    },
+    {
+      "epoch": 0.8499589263646526,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0017067561432890929,
+      "loss": 2.8184,
+      "step": 16668
+    },
+    {
+      "epoch": 0.850009919820758,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017062338897017055,
+      "loss": 2.7937,
+      "step": 16669
+    },
+    {
+      "epoch": 0.8500609132768634,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.001705711262158796,
+      "loss": 2.7868,
+      "step": 16670
+    },
+    {
+      "epoch": 0.8501119067329688,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001705188260982747,
+      "loss": 2.8108,
+      "step": 16671
+    },
+    {
+      "epoch": 0.8501629001890741,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0017046648864961719,
+      "loss": 2.7955,
+      "step": 16672
+    },
+    {
+      "epoch": 0.8502138936451795,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0017041411390219142,
+      "loss": 2.7721,
+      "step": 16673
+    },
+    {
+      "epoch": 0.850264887101285,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001703617018883048,
+      "loss": 2.7873,
+      "step": 16674
+    },
+    {
+      "epoch": 0.8503158805573904,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017030925264028773,
+      "loss": 2.8002,
+      "step": 16675
+    },
+    {
+      "epoch": 0.8503668740134958,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.001702567661904934,
+      "loss": 2.8052,
+      "step": 16676
+    },
+    {
+      "epoch": 0.8504178674696011,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0017020424257129822,
+      "loss": 2.7971,
+      "step": 16677
+    },
+    {
+      "epoch": 0.8504688609257065,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0017015168181510137,
+      "loss": 2.7972,
+      "step": 16678
+    },
+    {
+      "epoch": 0.8505198543818119,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017009908395432497,
+      "loss": 2.7904,
+      "step": 16679
+    },
+    {
+      "epoch": 0.8505708478379173,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017004644902141397,
+      "loss": 2.8184,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8506218412940227,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0016999377704883629,
+      "loss": 2.8079,
+      "step": 16681
+    },
+    {
+      "epoch": 0.8506728347501281,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0016994106806908266,
+      "loss": 2.7579,
+      "step": 16682
+    },
+    {
+      "epoch": 0.8507238282062335,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016988832211466659,
+      "loss": 2.7784,
+      "step": 16683
+    },
+    {
+      "epoch": 0.8507748216623389,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0016983553921812442,
+      "loss": 2.7845,
+      "step": 16684
+    },
+    {
+      "epoch": 0.8508258151184442,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0016978271941201536,
+      "loss": 2.782,
+      "step": 16685
+    },
+    {
+      "epoch": 0.8508768085745496,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0016972986272892128,
+      "loss": 2.7924,
+      "step": 16686
+    },
+    {
+      "epoch": 0.850927802030655,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001696769692014468,
+      "loss": 2.7881,
+      "step": 16687
+    },
+    {
+      "epoch": 0.8509787954867605,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0016962403886221938,
+      "loss": 2.8013,
+      "step": 16688
+    },
+    {
+      "epoch": 0.8510297889428659,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016957107174388909,
+      "loss": 2.8086,
+      "step": 16689
+    },
+    {
+      "epoch": 0.8510807823989712,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0016951806787912868,
+      "loss": 2.787,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8511317758550766,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0016946502730063363,
+      "loss": 2.7718,
+      "step": 16691
+    },
+    {
+      "epoch": 0.851182769311182,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0016941195004112198,
+      "loss": 2.8121,
+      "step": 16692
+    },
+    {
+      "epoch": 0.8512337627672873,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0016935883613333452,
+      "loss": 2.8155,
+      "step": 16693
+    },
+    {
+      "epoch": 0.8512847562233927,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0016930568561003454,
+      "loss": 2.8207,
+      "step": 16694
+    },
+    {
+      "epoch": 0.8513357496794982,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016925249850400796,
+      "loss": 2.7882,
+      "step": 16695
+    },
+    {
+      "epoch": 0.8513867431356036,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016919927484806325,
+      "loss": 2.7895,
+      "step": 16696
+    },
+    {
+      "epoch": 0.851437736591709,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016914601467503143,
+      "loss": 2.784,
+      "step": 16697
+    },
+    {
+      "epoch": 0.8514887300478143,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0016909271801776608,
+      "loss": 2.819,
+      "step": 16698
+    },
+    {
+      "epoch": 0.8515397235039197,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0016903938490914325,
+      "loss": 2.8243,
+      "step": 16699
+    },
+    {
+      "epoch": 0.8515907169600251,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0016898601538206145,
+      "loss": 2.8199,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8516417104161305,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0016893260946944171,
+      "loss": 2.8098,
+      "step": 16701
+    },
+    {
+      "epoch": 0.851692703872236,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016887916720422752,
+      "loss": 2.7483,
+      "step": 16702
+    },
+    {
+      "epoch": 0.8517436973283413,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016882568861938466,
+      "loss": 2.8014,
+      "step": 16703
+    },
+    {
+      "epoch": 0.8517946907844467,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0016877217374790146,
+      "loss": 2.788,
+      "step": 16704
+    },
+    {
+      "epoch": 0.8518456842405521,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001687186226227886,
+      "loss": 2.7808,
+      "step": 16705
+    },
+    {
+      "epoch": 0.8518966776966574,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0016866503527707911,
+      "loss": 2.79,
+      "step": 16706
+    },
+    {
+      "epoch": 0.8519476711527628,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001686114117438283,
+      "loss": 2.8004,
+      "step": 16707
+    },
+    {
+      "epoch": 0.8519986646088682,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0016855775205611389,
+      "loss": 2.7854,
+      "step": 16708
+    },
+    {
+      "epoch": 0.8520496580649737,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0016850405624703588,
+      "loss": 2.8142,
+      "step": 16709
+    },
+    {
+      "epoch": 0.852100651521079,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0016845032434971653,
+      "loss": 2.7765,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8521516449771844,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0016839655639730039,
+      "loss": 2.7826,
+      "step": 16711
+    },
+    {
+      "epoch": 0.8522026384332898,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0016834275242295423,
+      "loss": 2.7766,
+      "step": 16712
+    },
+    {
+      "epoch": 0.8522536318893952,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0016828891245986705,
+      "loss": 2.7878,
+      "step": 16713
+    },
+    {
+      "epoch": 0.8523046253455006,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0016823503654125002,
+      "loss": 2.8018,
+      "step": 16714
+    },
+    {
+      "epoch": 0.8523556188016059,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001681811247003365,
+      "loss": 2.79,
+      "step": 16715
+    },
+    {
+      "epoch": 0.8524066122577114,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0016812717697038207,
+      "loss": 2.7747,
+      "step": 16716
+    },
+    {
+      "epoch": 0.8524576057138168,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001680731933846644,
+      "loss": 2.8089,
+      "step": 16717
+    },
+    {
+      "epoch": 0.8525085991699222,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0016801917397648322,
+      "loss": 2.8095,
+      "step": 16718
+    },
+    {
+      "epoch": 0.8525595926260275,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0016796511877916049,
+      "loss": 2.791,
+      "step": 16719
+    },
+    {
+      "epoch": 0.8526105860821329,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001679110278260401,
+      "loss": 2.7933,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8526615795382383,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0016785690115048812,
+      "loss": 2.819,
+      "step": 16721
+    },
+    {
+      "epoch": 0.8527125729943437,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.001678027387858926,
+      "loss": 2.8158,
+      "step": 16722
+    },
+    {
+      "epoch": 0.8527635664504492,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0016774854076566357,
+      "loss": 2.8103,
+      "step": 16723
+    },
+    {
+      "epoch": 0.8528145599065545,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0016769430712323312,
+      "loss": 2.7842,
+      "step": 16724
+    },
+    {
+      "epoch": 0.8528655533626599,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.001676400378920553,
+      "loss": 2.8042,
+      "step": 16725
+    },
+    {
+      "epoch": 0.8529165468187653,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0016758573310560606,
+      "loss": 2.7891,
+      "step": 16726
+    },
+    {
+      "epoch": 0.8529675402748707,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0016753139279738335,
+      "loss": 2.7809,
+      "step": 16727
+    },
+    {
+      "epoch": 0.853018533730976,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0016747701700090697,
+      "loss": 2.7934,
+      "step": 16728
+    },
+    {
+      "epoch": 0.8530695271870814,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0016742260574971871,
+      "loss": 2.7846,
+      "step": 16729
+    },
+    {
+      "epoch": 0.8531205206431869,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016736815907738207,
+      "loss": 2.8038,
+      "step": 16730
+    },
+    {
+      "epoch": 0.8531715140992923,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0016731367701748256,
+      "loss": 2.7751,
+      "step": 16731
+    },
+    {
+      "epoch": 0.8532225075553976,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0016725915960362746,
+      "loss": 2.7892,
+      "step": 16732
+    },
+    {
+      "epoch": 0.853273501011503,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016720460686944578,
+      "loss": 2.7886,
+      "step": 16733
+    },
+    {
+      "epoch": 0.8533244944676084,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0016715001884858847,
+      "loss": 2.7798,
+      "step": 16734
+    },
+    {
+      "epoch": 0.8533754879237138,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.001670953955747281,
+      "loss": 2.7915,
+      "step": 16735
+    },
+    {
+      "epoch": 0.8534264813798191,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001670407370815591,
+      "loss": 2.7957,
+      "step": 16736
+    },
+    {
+      "epoch": 0.8534774748359246,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001669860434027975,
+      "loss": 2.8126,
+      "step": 16737
+    },
+    {
+      "epoch": 0.85352846829203,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.001669313145721812,
+      "loss": 2.8088,
+      "step": 16738
+    },
+    {
+      "epoch": 0.8535794617481354,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0016687655062346965,
+      "loss": 2.804,
+      "step": 16739
+    },
+    {
+      "epoch": 0.8536304552042407,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00166821751590444,
+      "loss": 2.8067,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8536814486603461,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0016676691750690704,
+      "loss": 2.7716,
+      "step": 16741
+    },
+    {
+      "epoch": 0.8537324421164515,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0016671204840668324,
+      "loss": 2.8028,
+      "step": 16742
+    },
+    {
+      "epoch": 0.8537834355725569,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016665714432361853,
+      "loss": 2.8242,
+      "step": 16743
+    },
+    {
+      "epoch": 0.8538344290286624,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0016660220529158061,
+      "loss": 2.8054,
+      "step": 16744
+    },
+    {
+      "epoch": 0.8538854224847677,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0016654723134445855,
+      "loss": 2.7518,
+      "step": 16745
+    },
+    {
+      "epoch": 0.8539364159408731,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0016649222251616305,
+      "loss": 2.8117,
+      "step": 16746
+    },
+    {
+      "epoch": 0.8539874093969785,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0016643717884062637,
+      "loss": 2.7889,
+      "step": 16747
+    },
+    {
+      "epoch": 0.8540384028530839,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0016638210035180216,
+      "loss": 2.7787,
+      "step": 16748
+    },
+    {
+      "epoch": 0.8540893963091892,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0016632698708366567,
+      "loss": 2.8153,
+      "step": 16749
+    },
+    {
+      "epoch": 0.8541403897652946,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016627183907021348,
+      "loss": 2.8,
+      "step": 16750
+    },
+    {
+      "epoch": 0.8541913832214001,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0016621665634546363,
+      "loss": 2.783,
+      "step": 16751
+    },
+    {
+      "epoch": 0.8542423766775055,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0016616143894345567,
+      "loss": 2.7802,
+      "step": 16752
+    },
+    {
+      "epoch": 0.8542933701336108,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0016610618689825041,
+      "loss": 2.7904,
+      "step": 16753
+    },
+    {
+      "epoch": 0.8543443635897162,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016605090024393013,
+      "loss": 2.8136,
+      "step": 16754
+    },
+    {
+      "epoch": 0.8543953570458216,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0016599557901459839,
+      "loss": 2.7757,
+      "step": 16755
+    },
+    {
+      "epoch": 0.854446350501927,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001659402232443801,
+      "loss": 2.7757,
+      "step": 16756
+    },
+    {
+      "epoch": 0.8544973439580323,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0016588483296742147,
+      "loss": 2.7885,
+      "step": 16757
+    },
+    {
+      "epoch": 0.8545483374141378,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0016582940821789004,
+      "loss": 2.7808,
+      "step": 16758
+    },
+    {
+      "epoch": 0.8545993308702432,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0016577394902997456,
+      "loss": 2.7869,
+      "step": 16759
+    },
+    {
+      "epoch": 0.8546503243263486,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001657184554378851,
+      "loss": 2.7869,
+      "step": 16760
+    },
+    {
+      "epoch": 0.854701317782454,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0016566292747585275,
+      "loss": 2.7996,
+      "step": 16761
+    },
+    {
+      "epoch": 0.8547523112385593,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0016560736517813011,
+      "loss": 2.771,
+      "step": 16762
+    },
+    {
+      "epoch": 0.8548033046946647,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0016555176857899069,
+      "loss": 2.8028,
+      "step": 16763
+    },
+    {
+      "epoch": 0.8548542981507701,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.001654961377127293,
+      "loss": 2.794,
+      "step": 16764
+    },
+    {
+      "epoch": 0.8549052916068756,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001654404726136619,
+      "loss": 2.7901,
+      "step": 16765
+    },
+    {
+      "epoch": 0.8549562850629809,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0016538477331612545,
+      "loss": 2.7691,
+      "step": 16766
+    },
+    {
+      "epoch": 0.8550072785190863,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001653290398544781,
+      "loss": 2.7929,
+      "step": 16767
+    },
+    {
+      "epoch": 0.8550582719751917,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0016527327226309905,
+      "loss": 2.7875,
+      "step": 16768
+    },
+    {
+      "epoch": 0.8551092654312971,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016521747057638854,
+      "loss": 2.7776,
+      "step": 16769
+    },
+    {
+      "epoch": 0.8551602588874024,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001651616348287679,
+      "loss": 2.7823,
+      "step": 16770
+    },
+    {
+      "epoch": 0.8552112523435078,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0016510576505467936,
+      "loss": 2.8111,
+      "step": 16771
+    },
+    {
+      "epoch": 0.8552622457996133,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016504986128858625,
+      "loss": 2.8048,
+      "step": 16772
+    },
+    {
+      "epoch": 0.8553132392557187,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0016499392356497283,
+      "loss": 2.8163,
+      "step": 16773
+    },
+    {
+      "epoch": 0.855364232711824,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0016493795191834422,
+      "loss": 2.7831,
+      "step": 16774
+    },
+    {
+      "epoch": 0.8554152261679294,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0016488194638322664,
+      "loss": 2.7563,
+      "step": 16775
+    },
+    {
+      "epoch": 0.8554662196240348,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0016482590699416707,
+      "loss": 2.8102,
+      "step": 16776
+    },
+    {
+      "epoch": 0.8555172130801402,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0016476983378573343,
+      "loss": 2.7965,
+      "step": 16777
+    },
+    {
+      "epoch": 0.8555682065362455,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.001647137267925145,
+      "loss": 2.8165,
+      "step": 16778
+    },
+    {
+      "epoch": 0.855619199992351,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0016465758604911986,
+      "loss": 2.8003,
+      "step": 16779
+    },
+    {
+      "epoch": 0.8556701934484564,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016460141159017996,
+      "loss": 2.7896,
+      "step": 16780
+    },
+    {
+      "epoch": 0.8557211869045618,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0016454520345034608,
+      "loss": 2.7814,
+      "step": 16781
+    },
+    {
+      "epoch": 0.8557721803606672,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0016448896166429013,
+      "loss": 2.8099,
+      "step": 16782
+    },
+    {
+      "epoch": 0.8558231738167725,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0016443268626670495,
+      "loss": 2.8136,
+      "step": 16783
+    },
+    {
+      "epoch": 0.8558741672728779,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016437637729230397,
+      "loss": 2.7716,
+      "step": 16784
+    },
+    {
+      "epoch": 0.8559251607289833,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0016432003477582142,
+      "loss": 2.7832,
+      "step": 16785
+    },
+    {
+      "epoch": 0.8559761541850888,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0016426365875201224,
+      "loss": 2.8149,
+      "step": 16786
+    },
+    {
+      "epoch": 0.8560271476411941,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016420724925565193,
+      "loss": 2.8036,
+      "step": 16787
+    },
+    {
+      "epoch": 0.8560781410972995,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016415080632153673,
+      "loss": 2.8203,
+      "step": 16788
+    },
+    {
+      "epoch": 0.8561291345534049,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0016409432998448345,
+      "loss": 2.7851,
+      "step": 16789
+    },
+    {
+      "epoch": 0.8561801280095103,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0016403782027932957,
+      "loss": 2.8086,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8562311214656156,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001639812772409331,
+      "loss": 2.7673,
+      "step": 16791
+    },
+    {
+      "epoch": 0.856282114921721,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0016392470090417265,
+      "loss": 2.8038,
+      "step": 16792
+    },
+    {
+      "epoch": 0.8563331083778265,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016386809130394726,
+      "loss": 2.7873,
+      "step": 16793
+    },
+    {
+      "epoch": 0.8563841018339319,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0016381144847517671,
+      "loss": 2.7831,
+      "step": 16794
+    },
+    {
+      "epoch": 0.8564350952900373,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0016375477245280103,
+      "loss": 2.7775,
+      "step": 16795
+    },
+    {
+      "epoch": 0.8564860887461426,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0016369806327178093,
+      "loss": 2.7981,
+      "step": 16796
+    },
+    {
+      "epoch": 0.856537082202248,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0016364132096709737,
+      "loss": 2.7854,
+      "step": 16797
+    },
+    {
+      "epoch": 0.8565880756583534,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0016358454557375195,
+      "loss": 2.7886,
+      "step": 16798
+    },
+    {
+      "epoch": 0.8566390691144588,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0016352773712676654,
+      "loss": 2.8,
+      "step": 16799
+    },
+    {
+      "epoch": 0.8566900625705642,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0016347089566118348,
+      "loss": 2.7809,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8567410560266696,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0016341402121206538,
+      "loss": 2.7674,
+      "step": 16801
+    },
+    {
+      "epoch": 0.856792049482775,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0016335711381449535,
+      "loss": 2.7769,
+      "step": 16802
+    },
+    {
+      "epoch": 0.8568430429388804,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0016330017350357664,
+      "loss": 2.806,
+      "step": 16803
+    },
+    {
+      "epoch": 0.8568940363949857,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0016324320031443295,
+      "loss": 2.7545,
+      "step": 16804
+    },
+    {
+      "epoch": 0.8569450298510911,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0016318619428220815,
+      "loss": 2.7961,
+      "step": 16805
+    },
+    {
+      "epoch": 0.8569960233071965,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0016312915544206652,
+      "loss": 2.7945,
+      "step": 16806
+    },
+    {
+      "epoch": 0.857047016763302,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001630720838291924,
+      "loss": 2.7986,
+      "step": 16807
+    },
+    {
+      "epoch": 0.8570980102194073,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0016301497947879048,
+      "loss": 2.8123,
+      "step": 16808
+    },
+    {
+      "epoch": 0.8571490036755127,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016295784242608552,
+      "loss": 2.8103,
+      "step": 16809
+    },
+    {
+      "epoch": 0.8571999971316181,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.001629006727063226,
+      "loss": 2.8131,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8572509905877235,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.001628434703547668,
+      "loss": 2.7797,
+      "step": 16811
+    },
+    {
+      "epoch": 0.8573019840438288,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0016278623540670351,
+      "loss": 2.7992,
+      "step": 16812
+    },
+    {
+      "epoch": 0.8573529774999342,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0016272896789743805,
+      "loss": 2.7828,
+      "step": 16813
+    },
+    {
+      "epoch": 0.8574039709560397,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0016267166786229593,
+      "loss": 2.798,
+      "step": 16814
+    },
+    {
+      "epoch": 0.8574549644121451,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0016261433533662265,
+      "loss": 2.801,
+      "step": 16815
+    },
+    {
+      "epoch": 0.8575059578682505,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0016255697035578377,
+      "loss": 2.7859,
+      "step": 16816
+    },
+    {
+      "epoch": 0.8575569513243558,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00162499572955165,
+      "loss": 2.7893,
+      "step": 16817
+    },
+    {
+      "epoch": 0.8576079447804612,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0016244214317017185,
+      "loss": 2.7926,
+      "step": 16818
+    },
+    {
+      "epoch": 0.8576589382365666,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0016238468103622992,
+      "loss": 2.795,
+      "step": 16819
+    },
+    {
+      "epoch": 0.857709931692672,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0016232718658878477,
+      "loss": 2.7958,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8577609251487774,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001622696598633018,
+      "loss": 2.7835,
+      "step": 16821
+    },
+    {
+      "epoch": 0.8578119186048828,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0016221210089526646,
+      "loss": 2.7707,
+      "step": 16822
+    },
+    {
+      "epoch": 0.8578629120609882,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0016215450972018395,
+      "loss": 2.8079,
+      "step": 16823
+    },
+    {
+      "epoch": 0.8579139055170936,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016209688637357942,
+      "loss": 2.7775,
+      "step": 16824
+    },
+    {
+      "epoch": 0.8579648989731989,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0016203923089099788,
+      "loss": 2.7712,
+      "step": 16825
+    },
+    {
+      "epoch": 0.8580158924293043,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0016198154330800407,
+      "loss": 2.7827,
+      "step": 16826
+    },
+    {
+      "epoch": 0.8580668858854097,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0016192382366018265,
+      "loss": 2.7987,
+      "step": 16827
+    },
+    {
+      "epoch": 0.8581178793415152,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0016186607198313795,
+      "loss": 2.7865,
+      "step": 16828
+    },
+    {
+      "epoch": 0.8581688727976206,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0016180828831249415,
+      "loss": 2.8183,
+      "step": 16829
+    },
+    {
+      "epoch": 0.8582198662537259,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0016175047268389509,
+      "loss": 2.8235,
+      "step": 16830
+    },
+    {
+      "epoch": 0.8582708597098313,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001616926251330043,
+      "loss": 2.7907,
+      "step": 16831
+    },
+    {
+      "epoch": 0.8583218531659367,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001616347456955052,
+      "loss": 2.7816,
+      "step": 16832
+    },
+    {
+      "epoch": 0.858372846622042,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0016157683440710061,
+      "loss": 2.8188,
+      "step": 16833
+    },
+    {
+      "epoch": 0.8584238400781474,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001615188913035132,
+      "loss": 2.793,
+      "step": 16834
+    },
+    {
+      "epoch": 0.8584748335342529,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001614609164204851,
+      "loss": 2.7812,
+      "step": 16835
+    },
+    {
+      "epoch": 0.8585258269903583,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0016140290979377821,
+      "loss": 2.7892,
+      "step": 16836
+    },
+    {
+      "epoch": 0.8585768204464637,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0016134487145917388,
+      "loss": 2.783,
+      "step": 16837
+    },
+    {
+      "epoch": 0.858627813902569,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0016128680145247314,
+      "loss": 2.7713,
+      "step": 16838
+    },
+    {
+      "epoch": 0.8586788073586744,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0016122869980949644,
+      "loss": 2.8089,
+      "step": 16839
+    },
+    {
+      "epoch": 0.8587298008147798,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0016117056656608376,
+      "loss": 2.7812,
+      "step": 16840
+    },
+    {
+      "epoch": 0.8587807942708852,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0016111240175809467,
+      "loss": 2.7732,
+      "step": 16841
+    },
+    {
+      "epoch": 0.8588317877269906,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0016105420542140809,
+      "loss": 2.81,
+      "step": 16842
+    },
+    {
+      "epoch": 0.858882781183096,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001609959775919225,
+      "loss": 2.7697,
+      "step": 16843
+    },
+    {
+      "epoch": 0.8589337746392014,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0016093771830555568,
+      "loss": 2.8029,
+      "step": 16844
+    },
+    {
+      "epoch": 0.8589847680953068,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0016087942759824495,
+      "loss": 2.8064,
+      "step": 16845
+    },
+    {
+      "epoch": 0.8590357615514121,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0016082110550594696,
+      "loss": 2.7834,
+      "step": 16846
+    },
+    {
+      "epoch": 0.8590867550075175,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0016076275206463762,
+      "loss": 2.7893,
+      "step": 16847
+    },
+    {
+      "epoch": 0.8591377484636229,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0016070436731031232,
+      "loss": 2.7956,
+      "step": 16848
+    },
+    {
+      "epoch": 0.8591887419197284,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0016064595127898572,
+      "loss": 2.7929,
+      "step": 16849
+    },
+    {
+      "epoch": 0.8592397353758338,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0016058750400669175,
+      "loss": 2.7884,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8592907288319391,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0016052902552948364,
+      "loss": 2.7922,
+      "step": 16851
+    },
+    {
+      "epoch": 0.8593417222880445,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001604705158834338,
+      "loss": 2.8017,
+      "step": 16852
+    },
+    {
+      "epoch": 0.8593927157441499,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0016041197510463397,
+      "loss": 2.7934,
+      "step": 16853
+    },
+    {
+      "epoch": 0.8594437092002553,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0016035340322919504,
+      "loss": 2.782,
+      "step": 16854
+    },
+    {
+      "epoch": 0.8594947026563606,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0016029480029324703,
+      "loss": 2.774,
+      "step": 16855
+    },
+    {
+      "epoch": 0.8595456961124661,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0016023616633293928,
+      "loss": 2.7848,
+      "step": 16856
+    },
+    {
+      "epoch": 0.8595966895685715,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0016017750138444006,
+      "loss": 2.7942,
+      "step": 16857
+    },
+    {
+      "epoch": 0.8596476830246769,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0016011880548393693,
+      "loss": 2.7642,
+      "step": 16858
+    },
+    {
+      "epoch": 0.8596986764807822,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0016006007866763642,
+      "loss": 2.7943,
+      "step": 16859
+    },
+    {
+      "epoch": 0.8597496699368876,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001600013209717642,
+      "loss": 2.7981,
+      "step": 16860
+    },
+    {
+      "epoch": 0.859800663392993,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015994253243256497,
+      "loss": 2.7874,
+      "step": 16861
+    },
+    {
+      "epoch": 0.8598516568490984,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0015988371308630247,
+      "loss": 2.7695,
+      "step": 16862
+    },
+    {
+      "epoch": 0.8599026503052039,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0015982486296925939,
+      "loss": 2.7552,
+      "step": 16863
+    },
+    {
+      "epoch": 0.8599536437613092,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001597659821177375,
+      "loss": 2.786,
+      "step": 16864
+    },
+    {
+      "epoch": 0.8600046372174146,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0015970707056805742,
+      "loss": 2.7671,
+      "step": 16865
+    },
+    {
+      "epoch": 0.86005563067352,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0015964812835655873,
+      "loss": 2.7723,
+      "step": 16866
+    },
+    {
+      "epoch": 0.8601066241296254,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.001595891555196,
+      "loss": 2.7931,
+      "step": 16867
+    },
+    {
+      "epoch": 0.8601576175857307,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001595301520935586,
+      "loss": 2.7783,
+      "step": 16868
+    },
+    {
+      "epoch": 0.8602086110418361,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0015947111811483088,
+      "loss": 2.7795,
+      "step": 16869
+    },
+    {
+      "epoch": 0.8602596044979416,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0015941205361983187,
+      "loss": 2.7729,
+      "step": 16870
+    },
+    {
+      "epoch": 0.860310597954047,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015935295864499557,
+      "loss": 2.7904,
+      "step": 16871
+    },
+    {
+      "epoch": 0.8603615914101523,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.001592938332267747,
+      "loss": 2.7632,
+      "step": 16872
+    },
+    {
+      "epoch": 0.8604125848662577,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0015923467740164084,
+      "loss": 2.7963,
+      "step": 16873
+    },
+    {
+      "epoch": 0.8604635783223631,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0015917549120608423,
+      "loss": 2.7756,
+      "step": 16874
+    },
+    {
+      "epoch": 0.8605145717784685,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0015911627467661389,
+      "loss": 2.8051,
+      "step": 16875
+    },
+    {
+      "epoch": 0.8605655652345738,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0015905702784975758,
+      "loss": 2.7687,
+      "step": 16876
+    },
+    {
+      "epoch": 0.8606165586906793,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0015899775076206168,
+      "loss": 2.785,
+      "step": 16877
+    },
+    {
+      "epoch": 0.8606675521467847,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0015893844345009132,
+      "loss": 2.7609,
+      "step": 16878
+    },
+    {
+      "epoch": 0.8607185456028901,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0015887910595043023,
+      "loss": 2.7962,
+      "step": 16879
+    },
+    {
+      "epoch": 0.8607695390589954,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0015881973829968068,
+      "loss": 2.7932,
+      "step": 16880
+    },
+    {
+      "epoch": 0.8608205325151008,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0015876034053446373,
+      "loss": 2.8063,
+      "step": 16881
+    },
+    {
+      "epoch": 0.8608715259712062,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0015870091269141883,
+      "loss": 2.7699,
+      "step": 16882
+    },
+    {
+      "epoch": 0.8609225194273116,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0015864145480720408,
+      "loss": 2.7976,
+      "step": 16883
+    },
+    {
+      "epoch": 0.8609735128834171,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0015858196691849612,
+      "loss": 2.7844,
+      "step": 16884
+    },
+    {
+      "epoch": 0.8610245063395224,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0015852244906199005,
+      "loss": 2.811,
+      "step": 16885
+    },
+    {
+      "epoch": 0.8610754997956278,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0015846290127439942,
+      "loss": 2.7767,
+      "step": 16886
+    },
+    {
+      "epoch": 0.8611264932517332,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0015840332359245638,
+      "loss": 2.7747,
+      "step": 16887
+    },
+    {
+      "epoch": 0.8611774867078386,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001583437160529114,
+      "loss": 2.7804,
+      "step": 16888
+    },
+    {
+      "epoch": 0.8612284801639439,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015828407869253338,
+      "loss": 2.7968,
+      "step": 16889
+    },
+    {
+      "epoch": 0.8612794736200493,
+      "grad_norm": 9.75,
+      "learning_rate": 0.001582244115481097,
+      "loss": 2.7883,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8613304670761548,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0015816471465644603,
+      "loss": 2.7877,
+      "step": 16891
+    },
+    {
+      "epoch": 0.8613814605322602,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0015810498805436635,
+      "loss": 2.7873,
+      "step": 16892
+    },
+    {
+      "epoch": 0.8614324539883655,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0015804523177871314,
+      "loss": 2.7851,
+      "step": 16893
+    },
+    {
+      "epoch": 0.8614834474444709,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0015798544586634697,
+      "loss": 2.7889,
+      "step": 16894
+    },
+    {
+      "epoch": 0.8615344409005763,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0015792563035414689,
+      "loss": 2.7803,
+      "step": 16895
+    },
+    {
+      "epoch": 0.8615854343566817,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0015786578527901,
+      "loss": 2.7992,
+      "step": 16896
+    },
+    {
+      "epoch": 0.861636427812787,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0015780591067785187,
+      "loss": 2.7531,
+      "step": 16897
+    },
+    {
+      "epoch": 0.8616874212688925,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0015774600658760608,
+      "loss": 2.7804,
+      "step": 16898
+    },
+    {
+      "epoch": 0.8617384147249979,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0015768607304522452,
+      "loss": 2.7745,
+      "step": 16899
+    },
+    {
+      "epoch": 0.8617894081811033,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0015762611008767716,
+      "loss": 2.7778,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8618404016372087,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0015756611775195222,
+      "loss": 2.7782,
+      "step": 16901
+    },
+    {
+      "epoch": 0.861891395093314,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00157506096075056,
+      "loss": 2.786,
+      "step": 16902
+    },
+    {
+      "epoch": 0.8619423885494194,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001574460450940128,
+      "loss": 2.7988,
+      "step": 16903
+    },
+    {
+      "epoch": 0.8619933820055248,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015738596484586515,
+      "loss": 2.7766,
+      "step": 16904
+    },
+    {
+      "epoch": 0.8620443754616303,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0015732585536767359,
+      "loss": 2.7817,
+      "step": 16905
+    },
+    {
+      "epoch": 0.8620953689177356,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0015726571669651664,
+      "loss": 2.7792,
+      "step": 16906
+    },
+    {
+      "epoch": 0.862146362373841,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0015720554886949078,
+      "loss": 2.7791,
+      "step": 16907
+    },
+    {
+      "epoch": 0.8621973558299464,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0015714535192371064,
+      "loss": 2.7681,
+      "step": 16908
+    },
+    {
+      "epoch": 0.8622483492860518,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001570851258963087,
+      "loss": 2.78,
+      "step": 16909
+    },
+    {
+      "epoch": 0.8622993427421571,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001570248708244354,
+      "loss": 2.7764,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8623503361982625,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001569645867452591,
+      "loss": 2.7871,
+      "step": 16911
+    },
+    {
+      "epoch": 0.862401329654368,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0015690427369596598,
+      "loss": 2.8049,
+      "step": 16912
+    },
+    {
+      "epoch": 0.8624523231104734,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0015684393171376029,
+      "loss": 2.772,
+      "step": 16913
+    },
+    {
+      "epoch": 0.8625033165665787,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015678356083586385,
+      "loss": 2.7608,
+      "step": 16914
+    },
+    {
+      "epoch": 0.8625543100226841,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0015672316109951656,
+      "loss": 2.7523,
+      "step": 16915
+    },
+    {
+      "epoch": 0.8626053034787895,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0015666273254197598,
+      "loss": 2.7596,
+      "step": 16916
+    },
+    {
+      "epoch": 0.8626562969348949,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001566022752005175,
+      "loss": 2.7895,
+      "step": 16917
+    },
+    {
+      "epoch": 0.8627072903910002,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0015654178911243416,
+      "loss": 2.783,
+      "step": 16918
+    },
+    {
+      "epoch": 0.8627582838471057,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001564812743150369,
+      "loss": 2.7638,
+      "step": 16919
+    },
+    {
+      "epoch": 0.8628092773032111,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0015642073084565426,
+      "loss": 2.7838,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8628602707593165,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0015636015874163254,
+      "loss": 2.7916,
+      "step": 16921
+    },
+    {
+      "epoch": 0.8629112642154219,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0015629955804033558,
+      "loss": 2.8156,
+      "step": 16922
+    },
+    {
+      "epoch": 0.8629622576715272,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00156238928779145,
+      "loss": 2.7548,
+      "step": 16923
+    },
+    {
+      "epoch": 0.8630132511276326,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0015617827099545993,
+      "loss": 2.7715,
+      "step": 16924
+    },
+    {
+      "epoch": 0.863064244583738,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001561175847266972,
+      "loss": 2.7842,
+      "step": 16925
+    },
+    {
+      "epoch": 0.8631152380398435,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0015605687001029106,
+      "loss": 2.7868,
+      "step": 16926
+    },
+    {
+      "epoch": 0.8631662314959488,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0015599612688369344,
+      "loss": 2.7824,
+      "step": 16927
+    },
+    {
+      "epoch": 0.8632172249520542,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015593535538437379,
+      "loss": 2.7871,
+      "step": 16928
+    },
+    {
+      "epoch": 0.8632682184081596,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00155874555549819,
+      "loss": 2.8042,
+      "step": 16929
+    },
+    {
+      "epoch": 0.863319211864265,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0015581372741753341,
+      "loss": 2.7755,
+      "step": 16930
+    },
+    {
+      "epoch": 0.8633702053203703,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0015575287102503896,
+      "loss": 2.7839,
+      "step": 16931
+    },
+    {
+      "epoch": 0.8634211987764757,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0015569198640987484,
+      "loss": 2.7826,
+      "step": 16932
+    },
+    {
+      "epoch": 0.8634721922325812,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0015563107360959782,
+      "loss": 2.7835,
+      "step": 16933
+    },
+    {
+      "epoch": 0.8635231856886866,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0015557013266178192,
+      "loss": 2.7789,
+      "step": 16934
+    },
+    {
+      "epoch": 0.863574179144792,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0015550916360401864,
+      "loss": 2.779,
+      "step": 16935
+    },
+    {
+      "epoch": 0.8636251726008973,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0015544816647391664,
+      "loss": 2.7739,
+      "step": 16936
+    },
+    {
+      "epoch": 0.8636761660570027,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0015538714130910217,
+      "loss": 2.7858,
+      "step": 16937
+    },
+    {
+      "epoch": 0.8637271595131081,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0015532608814721848,
+      "loss": 2.7387,
+      "step": 16938
+    },
+    {
+      "epoch": 0.8637781529692135,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.001552650070259263,
+      "loss": 2.8026,
+      "step": 16939
+    },
+    {
+      "epoch": 0.8638291464253189,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0015520389798290353,
+      "loss": 2.7846,
+      "step": 16940
+    },
+    {
+      "epoch": 0.8638801398814243,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0015514276105584527,
+      "loss": 2.7738,
+      "step": 16941
+    },
+    {
+      "epoch": 0.8639311333375297,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001550815962824639,
+      "loss": 2.7984,
+      "step": 16942
+    },
+    {
+      "epoch": 0.8639821267936351,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0015502040370048886,
+      "loss": 2.7622,
+      "step": 16943
+    },
+    {
+      "epoch": 0.8640331202497404,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0015495918334766684,
+      "loss": 2.8087,
+      "step": 16944
+    },
+    {
+      "epoch": 0.8640841137058458,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0015489793526176166,
+      "loss": 2.7606,
+      "step": 16945
+    },
+    {
+      "epoch": 0.8641351071619512,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0015483665948055416,
+      "loss": 2.7636,
+      "step": 16946
+    },
+    {
+      "epoch": 0.8641861006180567,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001547753560418424,
+      "loss": 2.7943,
+      "step": 16947
+    },
+    {
+      "epoch": 0.864237094074162,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001547140249834413,
+      "loss": 2.7654,
+      "step": 16948
+    },
+    {
+      "epoch": 0.8642880875302674,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0015465266634318304,
+      "loss": 2.82,
+      "step": 16949
+    },
+    {
+      "epoch": 0.8643390809863728,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015459128015891668,
+      "loss": 2.7791,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8643900744424782,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.001545298664685083,
+      "loss": 2.7915,
+      "step": 16951
+    },
+    {
+      "epoch": 0.8644410678985835,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0015446842530984094,
+      "loss": 2.7886,
+      "step": 16952
+    },
+    {
+      "epoch": 0.8644920613546889,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015440695672081465,
+      "loss": 2.7819,
+      "step": 16953
+    },
+    {
+      "epoch": 0.8645430548107944,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0015434546073934622,
+      "loss": 2.7782,
+      "step": 16954
+    },
+    {
+      "epoch": 0.8645940482668998,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0015428393740336962,
+      "loss": 2.8011,
+      "step": 16955
+    },
+    {
+      "epoch": 0.8646450417230052,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0015422238675083541,
+      "loss": 2.7871,
+      "step": 16956
+    },
+    {
+      "epoch": 0.8646960351791105,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.001541608088197112,
+      "loss": 2.7867,
+      "step": 16957
+    },
+    {
+      "epoch": 0.8647470286352159,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0015409920364798128,
+      "loss": 2.7457,
+      "step": 16958
+    },
+    {
+      "epoch": 0.8647980220913213,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.001540375712736469,
+      "loss": 2.805,
+      "step": 16959
+    },
+    {
+      "epoch": 0.8648490155474267,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0015397591173472596,
+      "loss": 2.7589,
+      "step": 16960
+    },
+    {
+      "epoch": 0.8649000090035321,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0015391422506925317,
+      "loss": 2.7563,
+      "step": 16961
+    },
+    {
+      "epoch": 0.8649510024596375,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001538525113152799,
+      "loss": 2.7775,
+      "step": 16962
+    },
+    {
+      "epoch": 0.8650019959157429,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0015379077051087442,
+      "loss": 2.7937,
+      "step": 16963
+    },
+    {
+      "epoch": 0.8650529893718483,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0015372900269412144,
+      "loss": 2.7876,
+      "step": 16964
+    },
+    {
+      "epoch": 0.8651039828279536,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0015366720790312254,
+      "loss": 2.8052,
+      "step": 16965
+    },
+    {
+      "epoch": 0.865154976284059,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0015360538617599582,
+      "loss": 2.766,
+      "step": 16966
+    },
+    {
+      "epoch": 0.8652059697401644,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0015354353755087603,
+      "loss": 2.7764,
+      "step": 16967
+    },
+    {
+      "epoch": 0.8652569631962699,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0015348166206591448,
+      "loss": 2.7582,
+      "step": 16968
+    },
+    {
+      "epoch": 0.8653079566523753,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0015341975975927914,
+      "loss": 2.7963,
+      "step": 16969
+    },
+    {
+      "epoch": 0.8653589501084806,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0015335783066915435,
+      "loss": 2.7894,
+      "step": 16970
+    },
+    {
+      "epoch": 0.865409943564586,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0015329587483374125,
+      "loss": 2.8074,
+      "step": 16971
+    },
+    {
+      "epoch": 0.8654609370206914,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0015323389229125717,
+      "loss": 2.7888,
+      "step": 16972
+    },
+    {
+      "epoch": 0.8655119304767968,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0015317188307993612,
+      "loss": 2.7721,
+      "step": 16973
+    },
+    {
+      "epoch": 0.8655629239329021,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.001531098472380285,
+      "loss": 2.7528,
+      "step": 16974
+    },
+    {
+      "epoch": 0.8656139173890076,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001530477848038011,
+      "loss": 2.7946,
+      "step": 16975
+    },
+    {
+      "epoch": 0.865664910845113,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001529856958155372,
+      "loss": 2.7661,
+      "step": 16976
+    },
+    {
+      "epoch": 0.8657159043012184,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0015292358031153632,
+      "loss": 2.763,
+      "step": 16977
+    },
+    {
+      "epoch": 0.8657668977573237,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0015286143833011451,
+      "loss": 2.8021,
+      "step": 16978
+    },
+    {
+      "epoch": 0.8658178912134291,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0015279926990960408,
+      "loss": 2.7862,
+      "step": 16979
+    },
+    {
+      "epoch": 0.8658688846695345,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0015273707508835352,
+      "loss": 2.8086,
+      "step": 16980
+    },
+    {
+      "epoch": 0.8659198781256399,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001526748539047278,
+      "loss": 2.7915,
+      "step": 16981
+    },
+    {
+      "epoch": 0.8659708715817453,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001526126063971081,
+      "loss": 2.805,
+      "step": 16982
+    },
+    {
+      "epoch": 0.8660218650378507,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0015255033260389175,
+      "loss": 2.7846,
+      "step": 16983
+    },
+    {
+      "epoch": 0.8660728584939561,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0015248803256349237,
+      "loss": 2.7731,
+      "step": 16984
+    },
+    {
+      "epoch": 0.8661238519500615,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.001524257063143398,
+      "loss": 2.7599,
+      "step": 16985
+    },
+    {
+      "epoch": 0.8661748454061668,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0015236335389487996,
+      "loss": 2.7608,
+      "step": 16986
+    },
+    {
+      "epoch": 0.8662258388622722,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015230097534357492,
+      "loss": 2.7586,
+      "step": 16987
+    },
+    {
+      "epoch": 0.8662768323183776,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0015223857069890296,
+      "loss": 2.7769,
+      "step": 16988
+    },
+    {
+      "epoch": 0.8663278257744831,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0015217613999935842,
+      "loss": 2.7733,
+      "step": 16989
+    },
+    {
+      "epoch": 0.8663788192305885,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0015211368328345162,
+      "loss": 2.7571,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8664298126866938,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0015205120058970902,
+      "loss": 2.7605,
+      "step": 16991
+    },
+    {
+      "epoch": 0.8664808061427992,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0015198869195667316,
+      "loss": 2.7919,
+      "step": 16992
+    },
+    {
+      "epoch": 0.8665317995989046,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0015192615742290242,
+      "loss": 2.7851,
+      "step": 16993
+    },
+    {
+      "epoch": 0.86658279305501,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0015186359702697126,
+      "loss": 2.7728,
+      "step": 16994
+    },
+    {
+      "epoch": 0.8666337865111153,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0015180101080747009,
+      "loss": 2.7976,
+      "step": 16995
+    },
+    {
+      "epoch": 0.8666847799672208,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0015173839880300525,
+      "loss": 2.7768,
+      "step": 16996
+    },
+    {
+      "epoch": 0.8667357734233262,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0015167576105219892,
+      "loss": 2.7739,
+      "step": 16997
+    },
+    {
+      "epoch": 0.8667867668794316,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015161309759368926,
+      "loss": 2.7892,
+      "step": 16998
+    },
+    {
+      "epoch": 0.8668377603355369,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0015155040846613023,
+      "loss": 2.7974,
+      "step": 16999
+    },
+    {
+      "epoch": 0.8668887537916423,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0015148769370819163,
+      "loss": 2.7837,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8669397472477477,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0015142495335855905,
+      "loss": 2.7749,
+      "step": 17001
+    },
+    {
+      "epoch": 0.8669907407038531,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0015136218745593394,
+      "loss": 2.7961,
+      "step": 17002
+    },
+    {
+      "epoch": 0.8670417341599586,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0015129939603903337,
+      "loss": 2.772,
+      "step": 17003
+    },
+    {
+      "epoch": 0.8670927276160639,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0015123657914659038,
+      "loss": 2.7863,
+      "step": 17004
+    },
+    {
+      "epoch": 0.8671437210721693,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0015117373681735344,
+      "loss": 2.7821,
+      "step": 17005
+    },
+    {
+      "epoch": 0.8671947145282747,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0015111086909008697,
+      "loss": 2.7935,
+      "step": 17006
+    },
+    {
+      "epoch": 0.86724570798438,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001510479760035709,
+      "loss": 2.7795,
+      "step": 17007
+    },
+    {
+      "epoch": 0.8672967014404854,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0015098505759660083,
+      "loss": 2.7868,
+      "step": 17008
+    },
+    {
+      "epoch": 0.8673476948965908,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0015092211390798796,
+      "loss": 2.7929,
+      "step": 17009
+    },
+    {
+      "epoch": 0.8673986883526963,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001508591449765592,
+      "loss": 2.7878,
+      "step": 17010
+    },
+    {
+      "epoch": 0.8674496818088017,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015079615084115695,
+      "loss": 2.7675,
+      "step": 17011
+    },
+    {
+      "epoch": 0.867500675264907,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001507331315406391,
+      "loss": 2.7873,
+      "step": 17012
+    },
+    {
+      "epoch": 0.8675516687210124,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0015067008711387915,
+      "loss": 2.7764,
+      "step": 17013
+    },
+    {
+      "epoch": 0.8676026621771178,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0015060701759976607,
+      "loss": 2.7611,
+      "step": 17014
+    },
+    {
+      "epoch": 0.8676536556332232,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001505439230372043,
+      "loss": 2.7992,
+      "step": 17015
+    },
+    {
+      "epoch": 0.8677046490893285,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0015048080346511373,
+      "loss": 2.7656,
+      "step": 17016
+    },
+    {
+      "epoch": 0.867755642545434,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0015041765892242966,
+      "loss": 2.7628,
+      "step": 17017
+    },
+    {
+      "epoch": 0.8678066360015394,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0015035448944810293,
+      "loss": 2.7756,
+      "step": 17018
+    },
+    {
+      "epoch": 0.8678576294576448,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001502912950810995,
+      "loss": 2.7706,
+      "step": 17019
+    },
+    {
+      "epoch": 0.8679086229137501,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0015022807586040092,
+      "loss": 2.7835,
+      "step": 17020
+    },
+    {
+      "epoch": 0.8679596163698555,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0015016483182500393,
+      "loss": 2.8211,
+      "step": 17021
+    },
+    {
+      "epoch": 0.8680106098259609,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0015010156301392067,
+      "loss": 2.7862,
+      "step": 17022
+    },
+    {
+      "epoch": 0.8680616032820663,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0015003826946617849,
+      "loss": 2.7364,
+      "step": 17023
+    },
+    {
+      "epoch": 0.8681125967381718,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0014997495122082009,
+      "loss": 2.7904,
+      "step": 17024
+    },
+    {
+      "epoch": 0.8681635901942771,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0014991160831690326,
+      "loss": 2.7878,
+      "step": 17025
+    },
+    {
+      "epoch": 0.8682145836503825,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014984824079350117,
+      "loss": 2.7593,
+      "step": 17026
+    },
+    {
+      "epoch": 0.8682655771064879,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0014978484868970199,
+      "loss": 2.7776,
+      "step": 17027
+    },
+    {
+      "epoch": 0.8683165705625933,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0014972143204460924,
+      "loss": 2.804,
+      "step": 17028
+    },
+    {
+      "epoch": 0.8683675640186986,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0014965799089734146,
+      "loss": 2.7958,
+      "step": 17029
+    },
+    {
+      "epoch": 0.868418557474804,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0014959452528703237,
+      "loss": 2.7905,
+      "step": 17030
+    },
+    {
+      "epoch": 0.8684695509309095,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.001495310352528307,
+      "loss": 2.7655,
+      "step": 17031
+    },
+    {
+      "epoch": 0.8685205443870149,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0014946752083390032,
+      "loss": 2.7983,
+      "step": 17032
+    },
+    {
+      "epoch": 0.8685715378431202,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0014940398206942011,
+      "loss": 2.7875,
+      "step": 17033
+    },
+    {
+      "epoch": 0.8686225312992256,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014934041899858398,
+      "loss": 2.7727,
+      "step": 17034
+    },
+    {
+      "epoch": 0.868673524755331,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0014927683166060077,
+      "loss": 2.7953,
+      "step": 17035
+    },
+    {
+      "epoch": 0.8687245182114364,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0014921322009469446,
+      "loss": 2.7678,
+      "step": 17036
+    },
+    {
+      "epoch": 0.8687755116675417,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0014914958434010374,
+      "loss": 2.7774,
+      "step": 17037
+    },
+    {
+      "epoch": 0.8688265051236472,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0014908592443608243,
+      "loss": 2.796,
+      "step": 17038
+    },
+    {
+      "epoch": 0.8688774985797526,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0014902224042189908,
+      "loss": 2.783,
+      "step": 17039
+    },
+    {
+      "epoch": 0.868928492035858,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0014895853233683727,
+      "loss": 2.7844,
+      "step": 17040
+    },
+    {
+      "epoch": 0.8689794854919634,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0014889480022019527,
+      "loss": 2.7685,
+      "step": 17041
+    },
+    {
+      "epoch": 0.8690304789480687,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001488310441112863,
+      "loss": 2.7798,
+      "step": 17042
+    },
+    {
+      "epoch": 0.8690814724041741,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0014876726404943834,
+      "loss": 2.744,
+      "step": 17043
+    },
+    {
+      "epoch": 0.8691324658602795,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0014870346007399412,
+      "loss": 2.7643,
+      "step": 17044
+    },
+    {
+      "epoch": 0.869183459316385,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001486396322243111,
+      "loss": 2.7774,
+      "step": 17045
+    },
+    {
+      "epoch": 0.8692344527724903,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0014857578053976154,
+      "loss": 2.7864,
+      "step": 17046
+    },
+    {
+      "epoch": 0.8692854462285957,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.001485119050597324,
+      "loss": 2.8105,
+      "step": 17047
+    },
+    {
+      "epoch": 0.8693364396847011,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0014844800582362525,
+      "loss": 2.7883,
+      "step": 17048
+    },
+    {
+      "epoch": 0.8693874331408065,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001483840828708563,
+      "loss": 2.7684,
+      "step": 17049
+    },
+    {
+      "epoch": 0.8694384265969118,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0014832013624085654,
+      "loss": 2.7661,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8694894200530172,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001482561659730714,
+      "loss": 2.7757,
+      "step": 17051
+    },
+    {
+      "epoch": 0.8695404135091227,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0014819217210696096,
+      "loss": 2.7715,
+      "step": 17052
+    },
+    {
+      "epoch": 0.8695914069652281,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014812815468199987,
+      "loss": 2.787,
+      "step": 17053
+    },
+    {
+      "epoch": 0.8696424004213335,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0014806411373767727,
+      "loss": 2.746,
+      "step": 17054
+    },
+    {
+      "epoch": 0.8696933938774388,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014800004931349688,
+      "loss": 2.7787,
+      "step": 17055
+    },
+    {
+      "epoch": 0.8697443873335442,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0014793596144897683,
+      "loss": 2.7763,
+      "step": 17056
+    },
+    {
+      "epoch": 0.8697953807896496,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0014787185018364968,
+      "loss": 2.7901,
+      "step": 17057
+    },
+    {
+      "epoch": 0.869846374245755,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0014780771555706258,
+      "loss": 2.7622,
+      "step": 17058
+    },
+    {
+      "epoch": 0.8698973677018604,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014774355760877689,
+      "loss": 2.7844,
+      "step": 17059
+    },
+    {
+      "epoch": 0.8699483611579658,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0014767937637836853,
+      "loss": 2.7801,
+      "step": 17060
+    },
+    {
+      "epoch": 0.8699993546140712,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0014761517190542768,
+      "loss": 2.7742,
+      "step": 17061
+    },
+    {
+      "epoch": 0.8700503480701766,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001475509442295589,
+      "loss": 2.8004,
+      "step": 17062
+    },
+    {
+      "epoch": 0.8701013415262819,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0014748669339038098,
+      "loss": 2.7727,
+      "step": 17063
+    },
+    {
+      "epoch": 0.8701523349823873,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0014742241942752718,
+      "loss": 2.7633,
+      "step": 17064
+    },
+    {
+      "epoch": 0.8702033284384927,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.001473581223806448,
+      "loss": 2.7934,
+      "step": 17065
+    },
+    {
+      "epoch": 0.8702543218945982,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0014729380228939559,
+      "loss": 2.793,
+      "step": 17066
+    },
+    {
+      "epoch": 0.8703053153507035,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001472294591934553,
+      "loss": 2.7889,
+      "step": 17067
+    },
+    {
+      "epoch": 0.8703563088068089,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0014716509313251408,
+      "loss": 2.8192,
+      "step": 17068
+    },
+    {
+      "epoch": 0.8704073022629143,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0014710070414627615,
+      "loss": 2.759,
+      "step": 17069
+    },
+    {
+      "epoch": 0.8704582957190197,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001470362922744598,
+      "loss": 2.7832,
+      "step": 17070
+    },
+    {
+      "epoch": 0.870509289175125,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0014697185755679753,
+      "loss": 2.7798,
+      "step": 17071
+    },
+    {
+      "epoch": 0.8705602826312304,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0014690740003303596,
+      "loss": 2.799,
+      "step": 17072
+    },
+    {
+      "epoch": 0.8706112760873359,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001468429197429357,
+      "loss": 2.7781,
+      "step": 17073
+    },
+    {
+      "epoch": 0.8706622695434413,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0014677841672627143,
+      "loss": 2.7499,
+      "step": 17074
+    },
+    {
+      "epoch": 0.8707132629995467,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0014671389102283183,
+      "loss": 2.7616,
+      "step": 17075
+    },
+    {
+      "epoch": 0.870764256455652,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0014664934267241965,
+      "loss": 2.7909,
+      "step": 17076
+    },
+    {
+      "epoch": 0.8708152499117574,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001465847717148515,
+      "loss": 2.7854,
+      "step": 17077
+    },
+    {
+      "epoch": 0.8708662433678628,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00146520178189958,
+      "loss": 2.7834,
+      "step": 17078
+    },
+    {
+      "epoch": 0.8709172368239682,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001464555621375837,
+      "loss": 2.7707,
+      "step": 17079
+    },
+    {
+      "epoch": 0.8709682302800736,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0014639092359758706,
+      "loss": 2.7801,
+      "step": 17080
+    },
+    {
+      "epoch": 0.871019223736179,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001463262626098403,
+      "loss": 2.7945,
+      "step": 17081
+    },
+    {
+      "epoch": 0.8710702171922844,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0014626157921422964,
+      "loss": 2.8086,
+      "step": 17082
+    },
+    {
+      "epoch": 0.8711212106483898,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0014619687345065499,
+      "loss": 2.7528,
+      "step": 17083
+    },
+    {
+      "epoch": 0.8711722041044951,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0014613214535903018,
+      "loss": 2.7682,
+      "step": 17084
+    },
+    {
+      "epoch": 0.8712231975606005,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0014606739497928269,
+      "loss": 2.7868,
+      "step": 17085
+    },
+    {
+      "epoch": 0.8712741910167059,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001460026223513538,
+      "loss": 2.7719,
+      "step": 17086
+    },
+    {
+      "epoch": 0.8713251844728114,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0014593782751519862,
+      "loss": 2.774,
+      "step": 17087
+    },
+    {
+      "epoch": 0.8713761779289168,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0014587301051078576,
+      "loss": 2.7655,
+      "step": 17088
+    },
+    {
+      "epoch": 0.8714271713850221,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014580817137809761,
+      "loss": 2.7965,
+      "step": 17089
+    },
+    {
+      "epoch": 0.8714781648411275,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0014574331015713025,
+      "loss": 2.757,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8715291582972329,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0014567842688789335,
+      "loss": 2.7863,
+      "step": 17091
+    },
+    {
+      "epoch": 0.8715801517533383,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001456135216104101,
+      "loss": 2.8018,
+      "step": 17092
+    },
+    {
+      "epoch": 0.8716311452094436,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0014554859436471737,
+      "loss": 2.8036,
+      "step": 17093
+    },
+    {
+      "epoch": 0.8716821386655491,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0014548364519086558,
+      "loss": 2.7835,
+      "step": 17094
+    },
+    {
+      "epoch": 0.8717331321216545,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0014541867412891862,
+      "loss": 2.7811,
+      "step": 17095
+    },
+    {
+      "epoch": 0.8717841255777599,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014535368121895392,
+      "loss": 2.7525,
+      "step": 17096
+    },
+    {
+      "epoch": 0.8718351190338652,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0014528866650106232,
+      "loss": 2.7858,
+      "step": 17097
+    },
+    {
+      "epoch": 0.8718861124899706,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0014522363001534823,
+      "loss": 2.7484,
+      "step": 17098
+    },
+    {
+      "epoch": 0.871937105946076,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0014515857180192936,
+      "loss": 2.753,
+      "step": 17099
+    },
+    {
+      "epoch": 0.8719880994021814,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0014509349190093696,
+      "loss": 2.7501,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8720390928582868,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0014502839035251552,
+      "loss": 2.7328,
+      "step": 17101
+    },
+    {
+      "epoch": 0.8720900863143922,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0014496326719682295,
+      "loss": 2.7603,
+      "step": 17102
+    },
+    {
+      "epoch": 0.8721410797704976,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0014489812247403052,
+      "loss": 2.785,
+      "step": 17103
+    },
+    {
+      "epoch": 0.872192073226603,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0014483295622432274,
+      "loss": 2.7655,
+      "step": 17104
+    },
+    {
+      "epoch": 0.8722430666827083,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0014476776848789746,
+      "loss": 2.7668,
+      "step": 17105
+    },
+    {
+      "epoch": 0.8722940601388137,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001447025593049657,
+      "loss": 2.7834,
+      "step": 17106
+    },
+    {
+      "epoch": 0.8723450535949191,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0014463732871575177,
+      "loss": 2.7806,
+      "step": 17107
+    },
+    {
+      "epoch": 0.8723960470510246,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0014457207676049325,
+      "loss": 2.7849,
+      "step": 17108
+    },
+    {
+      "epoch": 0.87244704050713,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0014450680347944072,
+      "loss": 2.7789,
+      "step": 17109
+    },
+    {
+      "epoch": 0.8724980339632353,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014444150891285807,
+      "loss": 2.7849,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8725490274193407,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001443761931010223,
+      "loss": 2.7656,
+      "step": 17111
+    },
+    {
+      "epoch": 0.8726000208754461,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0014431085608422342,
+      "loss": 2.7673,
+      "step": 17112
+    },
+    {
+      "epoch": 0.8726510143315515,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014424549790276462,
+      "loss": 2.7538,
+      "step": 17113
+    },
+    {
+      "epoch": 0.8727020077876568,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0014418011859696211,
+      "loss": 2.7699,
+      "step": 17114
+    },
+    {
+      "epoch": 0.8727530012437623,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0014411471820714513,
+      "loss": 2.7972,
+      "step": 17115
+    },
+    {
+      "epoch": 0.8728039946998677,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001440492967736559,
+      "loss": 2.8072,
+      "step": 17116
+    },
+    {
+      "epoch": 0.8728549881559731,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0014398385433684968,
+      "loss": 2.7871,
+      "step": 17117
+    },
+    {
+      "epoch": 0.8729059816120784,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0014391839093709465,
+      "loss": 2.774,
+      "step": 17118
+    },
+    {
+      "epoch": 0.8729569750681838,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001438529066147719,
+      "loss": 2.7974,
+      "step": 17119
+    },
+    {
+      "epoch": 0.8730079685242892,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0014378740141027548,
+      "loss": 2.7972,
+      "step": 17120
+    },
+    {
+      "epoch": 0.8730589619803946,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0014372187536401228,
+      "loss": 2.7702,
+      "step": 17121
+    },
+    {
+      "epoch": 0.8731099554365,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0014365632851640203,
+      "loss": 2.8064,
+      "step": 17122
+    },
+    {
+      "epoch": 0.8731609488926054,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0014359076090787737,
+      "loss": 2.7698,
+      "step": 17123
+    },
+    {
+      "epoch": 0.8732119423487108,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0014352517257888366,
+      "loss": 2.7833,
+      "step": 17124
+    },
+    {
+      "epoch": 0.8732629358048162,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001434595635698791,
+      "loss": 2.7937,
+      "step": 17125
+    },
+    {
+      "epoch": 0.8733139292609216,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001433939339213346,
+      "loss": 2.7909,
+      "step": 17126
+    },
+    {
+      "epoch": 0.8733649227170269,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0014332828367373382,
+      "loss": 2.7872,
+      "step": 17127
+    },
+    {
+      "epoch": 0.8734159161731323,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0014326261286757317,
+      "loss": 2.7926,
+      "step": 17128
+    },
+    {
+      "epoch": 0.8734669096292378,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0014319692154336169,
+      "loss": 2.7842,
+      "step": 17129
+    },
+    {
+      "epoch": 0.8735179030853432,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0014313120974162104,
+      "loss": 2.7689,
+      "step": 17130
+    },
+    {
+      "epoch": 0.8735688965414485,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014306547750288562,
+      "loss": 2.7867,
+      "step": 17131
+    },
+    {
+      "epoch": 0.8736198899975539,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0014299972486770231,
+      "loss": 2.74,
+      "step": 17132
+    },
+    {
+      "epoch": 0.8736708834536593,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0014293395187663072,
+      "loss": 2.7835,
+      "step": 17133
+    },
+    {
+      "epoch": 0.8737218769097647,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001428681585702429,
+      "loss": 2.7775,
+      "step": 17134
+    },
+    {
+      "epoch": 0.87377287036587,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.001428023449891235,
+      "loss": 2.7854,
+      "step": 17135
+    },
+    {
+      "epoch": 0.8738238638219755,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0014273651117386954,
+      "loss": 2.7583,
+      "step": 17136
+    },
+    {
+      "epoch": 0.8738748572780809,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0014267065716509076,
+      "loss": 2.7619,
+      "step": 17137
+    },
+    {
+      "epoch": 0.8739258507341863,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.001426047830034091,
+      "loss": 2.7554,
+      "step": 17138
+    },
+    {
+      "epoch": 0.8739768441902916,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0014253888872945916,
+      "loss": 2.7382,
+      "step": 17139
+    },
+    {
+      "epoch": 0.874027837646397,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0014247297438388774,
+      "loss": 2.7662,
+      "step": 17140
+    },
+    {
+      "epoch": 0.8740788311025024,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0014240704000735424,
+      "loss": 2.7653,
+      "step": 17141
+    },
+    {
+      "epoch": 0.8741298245586078,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0014234108564053017,
+      "loss": 2.782,
+      "step": 17142
+    },
+    {
+      "epoch": 0.8741808180147133,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0014227511132409957,
+      "loss": 2.7467,
+      "step": 17143
+    },
+    {
+      "epoch": 0.8742318114708186,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0014220911709875868,
+      "loss": 2.7425,
+      "step": 17144
+    },
+    {
+      "epoch": 0.874282804926924,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0014214310300521612,
+      "loss": 2.7762,
+      "step": 17145
+    },
+    {
+      "epoch": 0.8743337983830294,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0014207706908419256,
+      "loss": 2.7766,
+      "step": 17146
+    },
+    {
+      "epoch": 0.8743847918391348,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0014201101537642121,
+      "loss": 2.7755,
+      "step": 17147
+    },
+    {
+      "epoch": 0.8744357852952401,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0014194494192264718,
+      "loss": 2.7542,
+      "step": 17148
+    },
+    {
+      "epoch": 0.8744867787513455,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00141878848763628,
+      "loss": 2.7898,
+      "step": 17149
+    },
+    {
+      "epoch": 0.874537772207451,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0014181273594013315,
+      "loss": 2.7629,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8745887656635564,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0014174660349294441,
+      "loss": 2.76,
+      "step": 17151
+    },
+    {
+      "epoch": 0.8746397591196617,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0014168045146285562,
+      "loss": 2.7696,
+      "step": 17152
+    },
+    {
+      "epoch": 0.8746907525757671,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001416142798906726,
+      "loss": 2.786,
+      "step": 17153
+    },
+    {
+      "epoch": 0.8747417460318725,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001415480888172133,
+      "loss": 2.74,
+      "step": 17154
+    },
+    {
+      "epoch": 0.8747927394879779,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0014148187828330779,
+      "loss": 2.7886,
+      "step": 17155
+    },
+    {
+      "epoch": 0.8748437329440832,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00141415648329798,
+      "loss": 2.7693,
+      "step": 17156
+    },
+    {
+      "epoch": 0.8748947264001887,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0014134939899753786,
+      "loss": 2.7683,
+      "step": 17157
+    },
+    {
+      "epoch": 0.8749457198562941,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0014128313032739333,
+      "loss": 2.8025,
+      "step": 17158
+    },
+    {
+      "epoch": 0.8749967133123995,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0014121684236024226,
+      "loss": 2.7826,
+      "step": 17159
+    },
+    {
+      "epoch": 0.8750477067685049,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0014115053513697438,
+      "loss": 2.7742,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8750987002246102,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001410842086984913,
+      "loss": 2.7625,
+      "step": 17161
+    },
+    {
+      "epoch": 0.8751496936807156,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0014101786308570652,
+      "loss": 2.7621,
+      "step": 17162
+    },
+    {
+      "epoch": 0.875200687136821,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.001409514983395454,
+      "loss": 2.7513,
+      "step": 17163
+    },
+    {
+      "epoch": 0.8752516805929265,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0014088511450094494,
+      "loss": 2.7655,
+      "step": 17164
+    },
+    {
+      "epoch": 0.8753026740490318,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0014081871161085412,
+      "loss": 2.7837,
+      "step": 17165
+    },
+    {
+      "epoch": 0.8753536675051372,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0014075228971023352,
+      "loss": 2.7724,
+      "step": 17166
+    },
+    {
+      "epoch": 0.8754046609612426,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0014068584884005552,
+      "loss": 2.7586,
+      "step": 17167
+    },
+    {
+      "epoch": 0.875455654417348,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.001406193890413042,
+      "loss": 2.7711,
+      "step": 17168
+    },
+    {
+      "epoch": 0.8755066478734533,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0014055291035497532,
+      "loss": 2.7487,
+      "step": 17169
+    },
+    {
+      "epoch": 0.8755576413295587,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0014048641282207624,
+      "loss": 2.7726,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8756086347856642,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0014041989648362599,
+      "loss": 2.7489,
+      "step": 17171
+    },
+    {
+      "epoch": 0.8756596282417696,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0014035336138065518,
+      "loss": 2.7718,
+      "step": 17172
+    },
+    {
+      "epoch": 0.875710621697875,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0014028680755420605,
+      "loss": 2.7768,
+      "step": 17173
+    },
+    {
+      "epoch": 0.8757616151539803,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0014022023504533226,
+      "loss": 2.7688,
+      "step": 17174
+    },
+    {
+      "epoch": 0.8758126086100857,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0014015364389509915,
+      "loss": 2.7657,
+      "step": 17175
+    },
+    {
+      "epoch": 0.8758636020661911,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0014008703414458345,
+      "loss": 2.7364,
+      "step": 17176
+    },
+    {
+      "epoch": 0.8759145955222964,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0014002040583487346,
+      "loss": 2.7824,
+      "step": 17177
+    },
+    {
+      "epoch": 0.8759655889784019,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.001399537590070688,
+      "loss": 2.7677,
+      "step": 17178
+    },
+    {
+      "epoch": 0.8760165824345073,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0013988709370228061,
+      "loss": 2.7601,
+      "step": 17179
+    },
+    {
+      "epoch": 0.8760675758906127,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0013982040996163136,
+      "loss": 2.7668,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8761185693467181,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0013975370782625504,
+      "loss": 2.7805,
+      "step": 17181
+    },
+    {
+      "epoch": 0.8761695628028234,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0013968698733729677,
+      "loss": 2.7592,
+      "step": 17182
+    },
+    {
+      "epoch": 0.8762205562589288,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0013962024853591314,
+      "loss": 2.7572,
+      "step": 17183
+    },
+    {
+      "epoch": 0.8762715497150342,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0013955349146327197,
+      "loss": 2.752,
+      "step": 17184
+    },
+    {
+      "epoch": 0.8763225431711397,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0013948671616055243,
+      "loss": 2.7697,
+      "step": 17185
+    },
+    {
+      "epoch": 0.876373536627245,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001394199226689448,
+      "loss": 2.784,
+      "step": 17186
+    },
+    {
+      "epoch": 0.8764245300833504,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0013935311102965075,
+      "loss": 2.7837,
+      "step": 17187
+    },
+    {
+      "epoch": 0.8764755235394558,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0013928628128388302,
+      "loss": 2.777,
+      "step": 17188
+    },
+    {
+      "epoch": 0.8765265169955612,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0013921943347286558,
+      "loss": 2.7711,
+      "step": 17189
+    },
+    {
+      "epoch": 0.8765775104516665,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0013915256763783346,
+      "loss": 2.7794,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8766285039077719,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.001390856838200329,
+      "loss": 2.7696,
+      "step": 17191
+    },
+    {
+      "epoch": 0.8766794973638774,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0013901878206072121,
+      "loss": 2.7869,
+      "step": 17192
+    },
+    {
+      "epoch": 0.8767304908199828,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0013895186240116672,
+      "loss": 2.7585,
+      "step": 17193
+    },
+    {
+      "epoch": 0.8767814842760882,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0013888492488264887,
+      "loss": 2.7845,
+      "step": 17194
+    },
+    {
+      "epoch": 0.8768324777321935,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013881796954645807,
+      "loss": 2.7626,
+      "step": 17195
+    },
+    {
+      "epoch": 0.8768834711882989,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0013875099643389575,
+      "loss": 2.7664,
+      "step": 17196
+    },
+    {
+      "epoch": 0.8769344646444043,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.001386840055862743,
+      "loss": 2.797,
+      "step": 17197
+    },
+    {
+      "epoch": 0.8769854581005097,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0013861699704491696,
+      "loss": 2.7844,
+      "step": 17198
+    },
+    {
+      "epoch": 0.8770364515566151,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0013854997085115813,
+      "loss": 2.7456,
+      "step": 17199
+    },
+    {
+      "epoch": 0.8770874450127205,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0013848292704634274,
+      "loss": 2.743,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8771384384688259,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0013841586567182693,
+      "loss": 2.7757,
+      "step": 17201
+    },
+    {
+      "epoch": 0.8771894319249313,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0013834878676897745,
+      "loss": 2.7543,
+      "step": 17202
+    },
+    {
+      "epoch": 0.8772404253810366,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0013828169037917192,
+      "loss": 2.7642,
+      "step": 17203
+    },
+    {
+      "epoch": 0.877291418837142,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0013821457654379884,
+      "loss": 2.7854,
+      "step": 17204
+    },
+    {
+      "epoch": 0.8773424122932474,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0013814744530425734,
+      "loss": 2.7596,
+      "step": 17205
+    },
+    {
+      "epoch": 0.8773934057493529,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0013808029670195733,
+      "loss": 2.7599,
+      "step": 17206
+    },
+    {
+      "epoch": 0.8774443992054582,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.001380131307783195,
+      "loss": 2.7433,
+      "step": 17207
+    },
+    {
+      "epoch": 0.8774953926615636,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.001379459475747751,
+      "loss": 2.7583,
+      "step": 17208
+    },
+    {
+      "epoch": 0.877546386117669,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0013787874713276616,
+      "loss": 2.7739,
+      "step": 17209
+    },
+    {
+      "epoch": 0.8775973795737744,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0013781152949374528,
+      "loss": 2.7854,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8776483730298797,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0013774429469917565,
+      "loss": 2.7855,
+      "step": 17211
+    },
+    {
+      "epoch": 0.8776993664859851,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0013767704279053106,
+      "loss": 2.765,
+      "step": 17212
+    },
+    {
+      "epoch": 0.8777503599420906,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013760977380929589,
+      "loss": 2.801,
+      "step": 17213
+    },
+    {
+      "epoch": 0.877801353398196,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013754248779696508,
+      "loss": 2.7805,
+      "step": 17214
+    },
+    {
+      "epoch": 0.8778523468543014,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0013747518479504393,
+      "loss": 2.773,
+      "step": 17215
+    },
+    {
+      "epoch": 0.8779033403104067,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0013740786484504836,
+      "loss": 2.7939,
+      "step": 17216
+    },
+    {
+      "epoch": 0.8779543337665121,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0013734052798850468,
+      "loss": 2.7541,
+      "step": 17217
+    },
+    {
+      "epoch": 0.8780053272226175,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0013727317426694967,
+      "loss": 2.7855,
+      "step": 17218
+    },
+    {
+      "epoch": 0.8780563206787229,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0013720580372193045,
+      "loss": 2.7865,
+      "step": 17219
+    },
+    {
+      "epoch": 0.8781073141348283,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0013713841639500463,
+      "loss": 2.7613,
+      "step": 17220
+    },
+    {
+      "epoch": 0.8781583075909337,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0013707101232774003,
+      "loss": 2.7599,
+      "step": 17221
+    },
+    {
+      "epoch": 0.8782093010470391,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0013700359156171486,
+      "loss": 2.7642,
+      "step": 17222
+    },
+    {
+      "epoch": 0.8782602945031445,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001369361541385177,
+      "loss": 2.7701,
+      "step": 17223
+    },
+    {
+      "epoch": 0.8783112879592498,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001368687000997473,
+      "loss": 2.7706,
+      "step": 17224
+    },
+    {
+      "epoch": 0.8783622814153552,
+      "grad_norm": 0.5,
+      "learning_rate": 0.001368012294870127,
+      "loss": 2.7627,
+      "step": 17225
+    },
+    {
+      "epoch": 0.8784132748714606,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013673374234193314,
+      "loss": 2.7515,
+      "step": 17226
+    },
+    {
+      "epoch": 0.8784642683275661,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0013666623870613813,
+      "loss": 2.7322,
+      "step": 17227
+    },
+    {
+      "epoch": 0.8785152617836715,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001365987186212673,
+      "loss": 2.7919,
+      "step": 17228
+    },
+    {
+      "epoch": 0.8785662552397768,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0013653118212897041,
+      "loss": 2.7685,
+      "step": 17229
+    },
+    {
+      "epoch": 0.8786172486958822,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0013646362927090734,
+      "loss": 2.7815,
+      "step": 17230
+    },
+    {
+      "epoch": 0.8786682421519876,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0013639606008874817,
+      "loss": 2.7842,
+      "step": 17231
+    },
+    {
+      "epoch": 0.878719235608093,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0013632847462417285,
+      "loss": 2.7814,
+      "step": 17232
+    },
+    {
+      "epoch": 0.8787702290641983,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0013626087291887159,
+      "loss": 2.7931,
+      "step": 17233
+    },
+    {
+      "epoch": 0.8788212225203038,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0013619325501454446,
+      "loss": 2.7654,
+      "step": 17234
+    },
+    {
+      "epoch": 0.8788722159764092,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.001361256209529016,
+      "loss": 2.7407,
+      "step": 17235
+    },
+    {
+      "epoch": 0.8789232094325146,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0013605797077566313,
+      "loss": 2.8126,
+      "step": 17236
+    },
+    {
+      "epoch": 0.8789742028886199,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00135990304524559,
+      "loss": 2.7612,
+      "step": 17237
+    },
+    {
+      "epoch": 0.8790251963447253,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0013592262224132923,
+      "loss": 2.7867,
+      "step": 17238
+    },
+    {
+      "epoch": 0.8790761898008307,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0013585492396772363,
+      "loss": 2.8014,
+      "step": 17239
+    },
+    {
+      "epoch": 0.8791271832569361,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0013578720974550184,
+      "loss": 2.8074,
+      "step": 17240
+    },
+    {
+      "epoch": 0.8791781767130415,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0013571947961643347,
+      "loss": 2.7511,
+      "step": 17241
+    },
+    {
+      "epoch": 0.8792291701691469,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0013565173362229785,
+      "loss": 2.7499,
+      "step": 17242
+    },
+    {
+      "epoch": 0.8792801636252523,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001355839718048841,
+      "loss": 2.7752,
+      "step": 17243
+    },
+    {
+      "epoch": 0.8793311570813577,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013551619420599113,
+      "loss": 2.7633,
+      "step": 17244
+    },
+    {
+      "epoch": 0.879382150537463,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001354484008674275,
+      "loss": 2.789,
+      "step": 17245
+    },
+    {
+      "epoch": 0.8794331439935684,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0013538059183101163,
+      "loss": 2.7785,
+      "step": 17246
+    },
+    {
+      "epoch": 0.8794841374496738,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013531276713857154,
+      "loss": 2.7415,
+      "step": 17247
+    },
+    {
+      "epoch": 0.8795351309057793,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0013524492683194486,
+      "loss": 2.7749,
+      "step": 17248
+    },
+    {
+      "epoch": 0.8795861243618847,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0013517707095297893,
+      "loss": 2.8003,
+      "step": 17249
+    },
+    {
+      "epoch": 0.87963711781799,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0013510919954353065,
+      "loss": 2.7775,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8796881112740954,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.001350413126454665,
+      "loss": 2.7434,
+      "step": 17251
+    },
+    {
+      "epoch": 0.8797391047302008,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.001349734103006626,
+      "loss": 2.7589,
+      "step": 17252
+    },
+    {
+      "epoch": 0.8797900981863062,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0013490549255100442,
+      "loss": 2.744,
+      "step": 17253
+    },
+    {
+      "epoch": 0.8798410916424115,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0013483755943838713,
+      "loss": 2.7509,
+      "step": 17254
+    },
+    {
+      "epoch": 0.879892085098517,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0013476961100471524,
+      "loss": 2.7971,
+      "step": 17255
+    },
+    {
+      "epoch": 0.8799430785546224,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0013470164729190279,
+      "loss": 2.7888,
+      "step": 17256
+    },
+    {
+      "epoch": 0.8799940720107278,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0013463366834187318,
+      "loss": 2.7692,
+      "step": 17257
+    },
+    {
+      "epoch": 0.8800450654668331,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0013456567419655927,
+      "loss": 2.7486,
+      "step": 17258
+    },
+    {
+      "epoch": 0.8800960589229385,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0013449766489790325,
+      "loss": 2.7701,
+      "step": 17259
+    },
+    {
+      "epoch": 0.8801470523790439,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.001344296404878567,
+      "loss": 2.763,
+      "step": 17260
+    },
+    {
+      "epoch": 0.8801980458351493,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0013436160100838044,
+      "loss": 2.7751,
+      "step": 17261
+    },
+    {
+      "epoch": 0.8802490392912548,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001342935465014447,
+      "loss": 2.7737,
+      "step": 17262
+    },
+    {
+      "epoch": 0.8803000327473601,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013422547700902884,
+      "loss": 2.763,
+      "step": 17263
+    },
+    {
+      "epoch": 0.8803510262034655,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.001341573925731216,
+      "loss": 2.7608,
+      "step": 17264
+    },
+    {
+      "epoch": 0.8804020196595709,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0013408929323572092,
+      "loss": 2.7528,
+      "step": 17265
+    },
+    {
+      "epoch": 0.8804530131156763,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0013402117903883381,
+      "loss": 2.7564,
+      "step": 17266
+    },
+    {
+      "epoch": 0.8805040065717816,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0013395305002447658,
+      "loss": 2.7303,
+      "step": 17267
+    },
+    {
+      "epoch": 0.880555000027887,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.001338849062346746,
+      "loss": 2.7583,
+      "step": 17268
+    },
+    {
+      "epoch": 0.8806059934839925,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001338167477114624,
+      "loss": 2.7709,
+      "step": 17269
+    },
+    {
+      "epoch": 0.8806569869400979,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0013374857449688358,
+      "loss": 2.7735,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8807079803962032,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0013368038663299079,
+      "loss": 2.7604,
+      "step": 17271
+    },
+    {
+      "epoch": 0.8807589738523086,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0013361218416184577,
+      "loss": 2.7511,
+      "step": 17272
+    },
+    {
+      "epoch": 0.880809967308414,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0013354396712551917,
+      "loss": 2.7613,
+      "step": 17273
+    },
+    {
+      "epoch": 0.8808609607645194,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0013347573556609074,
+      "loss": 2.7828,
+      "step": 17274
+    },
+    {
+      "epoch": 0.8809119542206247,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0013340748952564909,
+      "loss": 2.7471,
+      "step": 17275
+    },
+    {
+      "epoch": 0.8809629476767302,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013333922904629188,
+      "loss": 2.7482,
+      "step": 17276
+    },
+    {
+      "epoch": 0.8810139411328356,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0013327095417012549,
+      "loss": 2.7578,
+      "step": 17277
+    },
+    {
+      "epoch": 0.881064934588941,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0013320266493926538,
+      "loss": 2.7254,
+      "step": 17278
+    },
+    {
+      "epoch": 0.8811159280450463,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0013313436139583577,
+      "loss": 2.7709,
+      "step": 17279
+    },
+    {
+      "epoch": 0.8811669215011517,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001330660435819697,
+      "loss": 2.7856,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8812179149572571,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0013299771153980902,
+      "loss": 2.7644,
+      "step": 17281
+    },
+    {
+      "epoch": 0.8812689084133625,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0013292936531150439,
+      "loss": 2.7595,
+      "step": 17282
+    },
+    {
+      "epoch": 0.881319901869468,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001328610049392152,
+      "loss": 2.7404,
+      "step": 17283
+    },
+    {
+      "epoch": 0.8813708953255733,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0013279263046510955,
+      "loss": 2.7761,
+      "step": 17284
+    },
+    {
+      "epoch": 0.8814218887816787,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013272424193136421,
+      "loss": 2.7441,
+      "step": 17285
+    },
+    {
+      "epoch": 0.8814728822377841,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0013265583938016474,
+      "loss": 2.7717,
+      "step": 17286
+    },
+    {
+      "epoch": 0.8815238756938895,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0013258742285370522,
+      "loss": 2.76,
+      "step": 17287
+    },
+    {
+      "epoch": 0.8815748691499948,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0013251899239418842,
+      "loss": 2.7728,
+      "step": 17288
+    },
+    {
+      "epoch": 0.8816258626061002,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0013245054804382568,
+      "loss": 2.765,
+      "step": 17289
+    },
+    {
+      "epoch": 0.8816768560622057,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0013238208984483696,
+      "loss": 2.7789,
+      "step": 17290
+    },
+    {
+      "epoch": 0.8817278495183111,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0013231361783945061,
+      "loss": 2.7453,
+      "step": 17291
+    },
+    {
+      "epoch": 0.8817788429744164,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0013224513206990366,
+      "loss": 2.7745,
+      "step": 17292
+    },
+    {
+      "epoch": 0.8818298364305218,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0013217663257844164,
+      "loss": 2.7667,
+      "step": 17293
+    },
+    {
+      "epoch": 0.8818808298866272,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.001321081194073184,
+      "loss": 2.791,
+      "step": 17294
+    },
+    {
+      "epoch": 0.8819318233427326,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0013203959259879628,
+      "loss": 2.7589,
+      "step": 17295
+    },
+    {
+      "epoch": 0.8819828167988379,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0013197105219514614,
+      "loss": 2.7732,
+      "step": 17296
+    },
+    {
+      "epoch": 0.8820338102549434,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.001319024982386471,
+      "loss": 2.7734,
+      "step": 17297
+    },
+    {
+      "epoch": 0.8820848037110488,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0013183393077158666,
+      "loss": 2.7494,
+      "step": 17298
+    },
+    {
+      "epoch": 0.8821357971671542,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0013176534983626075,
+      "loss": 2.7714,
+      "step": 17299
+    },
+    {
+      "epoch": 0.8821867906232596,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001316967554749735,
+      "loss": 2.7361,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8822377840793649,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0013162814773003734,
+      "loss": 2.751,
+      "step": 17301
+    },
+    {
+      "epoch": 0.8822887775354703,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0013155952664377297,
+      "loss": 2.7588,
+      "step": 17302
+    },
+    {
+      "epoch": 0.8823397709915757,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0013149089225850937,
+      "loss": 2.768,
+      "step": 17303
+    },
+    {
+      "epoch": 0.8823907644476812,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0013142224461658364,
+      "loss": 2.768,
+      "step": 17304
+    },
+    {
+      "epoch": 0.8824417579037865,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.001313535837603411,
+      "loss": 2.7577,
+      "step": 17305
+    },
+    {
+      "epoch": 0.8824927513598919,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0013128490973213522,
+      "loss": 2.7642,
+      "step": 17306
+    },
+    {
+      "epoch": 0.8825437448159973,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.001312162225743276,
+      "loss": 2.7236,
+      "step": 17307
+    },
+    {
+      "epoch": 0.8825947382721027,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0013114752232928793,
+      "loss": 2.7556,
+      "step": 17308
+    },
+    {
+      "epoch": 0.882645731728208,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0013107880903939395,
+      "loss": 2.771,
+      "step": 17309
+    },
+    {
+      "epoch": 0.8826967251843134,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001310100827470315,
+      "loss": 2.7619,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8827477186404189,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001309413434945944,
+      "loss": 2.759,
+      "step": 17311
+    },
+    {
+      "epoch": 0.8827987120965243,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0013087259132448443,
+      "loss": 2.7593,
+      "step": 17312
+    },
+    {
+      "epoch": 0.8828497055526296,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0013080382627911145,
+      "loss": 2.7509,
+      "step": 17313
+    },
+    {
+      "epoch": 0.882900699008735,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0013073504840089316,
+      "loss": 2.7614,
+      "step": 17314
+    },
+    {
+      "epoch": 0.8829516924648404,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013066625773225523,
+      "loss": 2.7642,
+      "step": 17315
+    },
+    {
+      "epoch": 0.8830026859209458,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0013059745431563117,
+      "loss": 2.7753,
+      "step": 17316
+    },
+    {
+      "epoch": 0.8830536793770511,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.001305286381934624,
+      "loss": 2.7765,
+      "step": 17317
+    },
+    {
+      "epoch": 0.8831046728331566,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.001304598094081982,
+      "loss": 2.7606,
+      "step": 17318
+    },
+    {
+      "epoch": 0.883155666289262,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0013039096800229555,
+      "loss": 2.7934,
+      "step": 17319
+    },
+    {
+      "epoch": 0.8832066597453674,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0013032211401821934,
+      "loss": 2.757,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8832576532014728,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0013025324749844215,
+      "loss": 2.7597,
+      "step": 17321
+    },
+    {
+      "epoch": 0.8833086466575781,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0013018436848544431,
+      "loss": 2.7832,
+      "step": 17322
+    },
+    {
+      "epoch": 0.8833596401136835,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0013011547702171384,
+      "loss": 2.7308,
+      "step": 17323
+    },
+    {
+      "epoch": 0.8834106335697889,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.001300465731497465,
+      "loss": 2.7668,
+      "step": 17324
+    },
+    {
+      "epoch": 0.8834616270258944,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0012997765691204564,
+      "loss": 2.7508,
+      "step": 17325
+    },
+    {
+      "epoch": 0.8835126204819997,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0012990872835112227,
+      "loss": 2.74,
+      "step": 17326
+    },
+    {
+      "epoch": 0.8835636139381051,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001298397875094949,
+      "loss": 2.7522,
+      "step": 17327
+    },
+    {
+      "epoch": 0.8836146073942105,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0012977083442968983,
+      "loss": 2.756,
+      "step": 17328
+    },
+    {
+      "epoch": 0.8836656008503159,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0012970186915424074,
+      "loss": 2.7779,
+      "step": 17329
+    },
+    {
+      "epoch": 0.8837165943064212,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0012963289172568883,
+      "loss": 2.7727,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8837675877625266,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0012956390218658291,
+      "loss": 2.7576,
+      "step": 17331
+    },
+    {
+      "epoch": 0.8838185812186321,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0012949490057947914,
+      "loss": 2.7484,
+      "step": 17332
+    },
+    {
+      "epoch": 0.8838695746747375,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0012942588694694124,
+      "loss": 2.7337,
+      "step": 17333
+    },
+    {
+      "epoch": 0.8839205681308429,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.001293568613315402,
+      "loss": 2.7265,
+      "step": 17334
+    },
+    {
+      "epoch": 0.8839715615869482,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0012928782377585458,
+      "loss": 2.7656,
+      "step": 17335
+    },
+    {
+      "epoch": 0.8840225550430536,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0012921877432247016,
+      "loss": 2.7651,
+      "step": 17336
+    },
+    {
+      "epoch": 0.884073548499159,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001291497130139801,
+      "loss": 2.7743,
+      "step": 17337
+    },
+    {
+      "epoch": 0.8841245419552644,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0012908063989298493,
+      "loss": 2.7746,
+      "step": 17338
+    },
+    {
+      "epoch": 0.8841755354113698,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0012901155500209237,
+      "loss": 2.7572,
+      "step": 17339
+    },
+    {
+      "epoch": 0.8842265288674752,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0012894245838391749,
+      "loss": 2.7671,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8842775223235806,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0012887335008108258,
+      "loss": 2.7765,
+      "step": 17341
+    },
+    {
+      "epoch": 0.884328515779686,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0012880423013621706,
+      "loss": 2.7828,
+      "step": 17342
+    },
+    {
+      "epoch": 0.8843795092357913,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.001287350985919576,
+      "loss": 2.7657,
+      "step": 17343
+    },
+    {
+      "epoch": 0.8844305026918967,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0012866595549094802,
+      "loss": 2.7384,
+      "step": 17344
+    },
+    {
+      "epoch": 0.8844814961480021,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0012859680087583922,
+      "loss": 2.7637,
+      "step": 17345
+    },
+    {
+      "epoch": 0.8845324896041076,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0012852763478928932,
+      "loss": 2.7784,
+      "step": 17346
+    },
+    {
+      "epoch": 0.884583483060213,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0012845845727396338,
+      "loss": 2.7729,
+      "step": 17347
+    },
+    {
+      "epoch": 0.8846344765163183,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0012838926837253357,
+      "loss": 2.7488,
+      "step": 17348
+    },
+    {
+      "epoch": 0.8846854699724237,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0012832006812767908,
+      "loss": 2.7354,
+      "step": 17349
+    },
+    {
+      "epoch": 0.8847364634285291,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0012825085658208612,
+      "loss": 2.7135,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8847874568846344,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0012818163377844784,
+      "loss": 2.7395,
+      "step": 17351
+    },
+    {
+      "epoch": 0.8848384503407398,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0012811239975946433,
+      "loss": 2.7888,
+      "step": 17352
+    },
+    {
+      "epoch": 0.8848894437968453,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0012804315456784263,
+      "loss": 2.7627,
+      "step": 17353
+    },
+    {
+      "epoch": 0.8849404372529507,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0012797389824629661,
+      "loss": 2.7816,
+      "step": 17354
+    },
+    {
+      "epoch": 0.8849914307090561,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0012790463083754713,
+      "loss": 2.739,
+      "step": 17355
+    },
+    {
+      "epoch": 0.8850424241651614,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0012783535238432176,
+      "loss": 2.7912,
+      "step": 17356
+    },
+    {
+      "epoch": 0.8850934176212668,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0012776606292935488,
+      "loss": 2.7685,
+      "step": 17357
+    },
+    {
+      "epoch": 0.8851444110773722,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001276967625153878,
+      "loss": 2.7729,
+      "step": 17358
+    },
+    {
+      "epoch": 0.8851954045334776,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0012762745118516845,
+      "loss": 2.7368,
+      "step": 17359
+    },
+    {
+      "epoch": 0.885246397989583,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0012755812898145155,
+      "loss": 2.7422,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8852973914456884,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0012748879594699852,
+      "loss": 2.8041,
+      "step": 17361
+    },
+    {
+      "epoch": 0.8853483849017938,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0012741945212457744,
+      "loss": 2.7881,
+      "step": 17362
+    },
+    {
+      "epoch": 0.8853993783578992,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0012735009755696308,
+      "loss": 2.7655,
+      "step": 17363
+    },
+    {
+      "epoch": 0.8854503718140045,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0012728073228693679,
+      "loss": 2.798,
+      "step": 17364
+    },
+    {
+      "epoch": 0.8855013652701099,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0012721135635728664,
+      "loss": 2.7669,
+      "step": 17365
+    },
+    {
+      "epoch": 0.8855523587262153,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0012714196981080711,
+      "loss": 2.7639,
+      "step": 17366
+    },
+    {
+      "epoch": 0.8856033521823208,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0012707257269029932,
+      "loss": 2.7411,
+      "step": 17367
+    },
+    {
+      "epoch": 0.8856543456384262,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001270031650385709,
+      "loss": 2.7736,
+      "step": 17368
+    },
+    {
+      "epoch": 0.8857053390945315,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0012693374689843602,
+      "loss": 2.7673,
+      "step": 17369
+    },
+    {
+      "epoch": 0.8857563325506369,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0012686431831271524,
+      "loss": 2.7736,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8858073260067423,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0012679487932423558,
+      "loss": 2.7586,
+      "step": 17371
+    },
+    {
+      "epoch": 0.8858583194628477,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0012672542997583052,
+      "loss": 2.7621,
+      "step": 17372
+    },
+    {
+      "epoch": 0.885909312918953,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0012665597031033993,
+      "loss": 2.7786,
+      "step": 17373
+    },
+    {
+      "epoch": 0.8859603063750585,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0012658650037061002,
+      "loss": 2.7717,
+      "step": 17374
+    },
+    {
+      "epoch": 0.8860112998311639,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0012651702019949329,
+      "loss": 2.7616,
+      "step": 17375
+    },
+    {
+      "epoch": 0.8860622932872693,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0012644752983984862,
+      "loss": 2.7604,
+      "step": 17376
+    },
+    {
+      "epoch": 0.8861132867433746,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.001263780293345412,
+      "loss": 2.7516,
+      "step": 17377
+    },
+    {
+      "epoch": 0.88616428019948,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0012630851872644235,
+      "loss": 2.7653,
+      "step": 17378
+    },
+    {
+      "epoch": 0.8862152736555854,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001262389980584298,
+      "loss": 2.7699,
+      "step": 17379
+    },
+    {
+      "epoch": 0.8862662671116908,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0012616946737338734,
+      "loss": 2.7704,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8863172605677963,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0012609992671420499,
+      "loss": 2.7711,
+      "step": 17381
+    },
+    {
+      "epoch": 0.8863682540239016,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0012603037612377891,
+      "loss": 2.76,
+      "step": 17382
+    },
+    {
+      "epoch": 0.886419247480007,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0012596081564501143,
+      "loss": 2.7469,
+      "step": 17383
+    },
+    {
+      "epoch": 0.8864702409361124,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0012589124532081095,
+      "loss": 2.7577,
+      "step": 17384
+    },
+    {
+      "epoch": 0.8865212343922177,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0012582166519409194,
+      "loss": 2.7431,
+      "step": 17385
+    },
+    {
+      "epoch": 0.8865722278483231,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0012575207530777486,
+      "loss": 2.754,
+      "step": 17386
+    },
+    {
+      "epoch": 0.8866232213044285,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0012568247570478631,
+      "loss": 2.739,
+      "step": 17387
+    },
+    {
+      "epoch": 0.886674214760534,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.001256128664280588,
+      "loss": 2.7699,
+      "step": 17388
+    },
+    {
+      "epoch": 0.8867252082166394,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0012554324752053084,
+      "loss": 2.7584,
+      "step": 17389
+    },
+    {
+      "epoch": 0.8867762016727447,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0012547361902514684,
+      "loss": 2.7623,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8868271951288501,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.001254039809848572,
+      "loss": 2.7429,
+      "step": 17391
+    },
+    {
+      "epoch": 0.8868781885849555,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0012533433344261817,
+      "loss": 2.7557,
+      "step": 17392
+    },
+    {
+      "epoch": 0.8869291820410609,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001252646764413918,
+      "loss": 2.7629,
+      "step": 17393
+    },
+    {
+      "epoch": 0.8869801754971662,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0012519501002414605,
+      "loss": 2.7642,
+      "step": 17394
+    },
+    {
+      "epoch": 0.8870311689532717,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0012512533423385467,
+      "loss": 2.7925,
+      "step": 17395
+    },
+    {
+      "epoch": 0.8870821624093771,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0012505564911349718,
+      "loss": 2.7614,
+      "step": 17396
+    },
+    {
+      "epoch": 0.8871331558654825,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.001249859547060589,
+      "loss": 2.7799,
+      "step": 17397
+    },
+    {
+      "epoch": 0.8871841493215878,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0012491625105453077,
+      "loss": 2.7742,
+      "step": 17398
+    },
+    {
+      "epoch": 0.8872351427776932,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0012484653820190959,
+      "loss": 2.778,
+      "step": 17399
+    },
+    {
+      "epoch": 0.8872861362337986,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0012477681619119767,
+      "loss": 2.77,
+      "step": 17400
+    },
+    {
+      "epoch": 0.887337129689904,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0012470708506540312,
+      "loss": 2.7451,
+      "step": 17401
+    },
+    {
+      "epoch": 0.8873881231460095,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0012463734486753953,
+      "loss": 2.7721,
+      "step": 17402
+    },
+    {
+      "epoch": 0.8874391166021148,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0012456759564062623,
+      "loss": 2.7501,
+      "step": 17403
+    },
+    {
+      "epoch": 0.8874901100582202,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0012449783742768798,
+      "loss": 2.7763,
+      "step": 17404
+    },
+    {
+      "epoch": 0.8875411035143256,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0012442807027175523,
+      "loss": 2.7362,
+      "step": 17405
+    },
+    {
+      "epoch": 0.887592096970431,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0012435829421586378,
+      "loss": 2.7519,
+      "step": 17406
+    },
+    {
+      "epoch": 0.8876430904265363,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0012428850930305507,
+      "loss": 2.7749,
+      "step": 17407
+    },
+    {
+      "epoch": 0.8876940838826417,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001242187155763759,
+      "loss": 2.7261,
+      "step": 17408
+    },
+    {
+      "epoch": 0.8877450773387472,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0012414891307887858,
+      "loss": 2.765,
+      "step": 17409
+    },
+    {
+      "epoch": 0.8877960707948526,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.001240791018536208,
+      "loss": 2.7445,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8878470642509579,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0012400928194366558,
+      "loss": 2.7458,
+      "step": 17411
+    },
+    {
+      "epoch": 0.8878980577070633,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0012393945339208138,
+      "loss": 2.766,
+      "step": 17412
+    },
+    {
+      "epoch": 0.8879490511631687,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0012386961624194202,
+      "loss": 2.7517,
+      "step": 17413
+    },
+    {
+      "epoch": 0.8880000446192741,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0012379977053632646,
+      "loss": 2.776,
+      "step": 17414
+    },
+    {
+      "epoch": 0.8880510380753794,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.001237299163183191,
+      "loss": 2.7769,
+      "step": 17415
+    },
+    {
+      "epoch": 0.8881020315314849,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.001236600536310095,
+      "loss": 2.7507,
+      "step": 17416
+    },
+    {
+      "epoch": 0.8881530249875903,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0012359018251749255,
+      "loss": 2.7673,
+      "step": 17417
+    },
+    {
+      "epoch": 0.8882040184436957,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0012352030302086815,
+      "loss": 2.7819,
+      "step": 17418
+    },
+    {
+      "epoch": 0.888255011899801,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0012345041518424156,
+      "loss": 2.7928,
+      "step": 17419
+    },
+    {
+      "epoch": 0.8883060053559064,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001233805190507231,
+      "loss": 2.7431,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8883569988120118,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0012331061466342818,
+      "loss": 2.7552,
+      "step": 17421
+    },
+    {
+      "epoch": 0.8884079922681172,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0012324070206547737,
+      "loss": 2.76,
+      "step": 17422
+    },
+    {
+      "epoch": 0.8884589857242227,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0012317078129999626,
+      "loss": 2.7291,
+      "step": 17423
+    },
+    {
+      "epoch": 0.888509979180328,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0012310085241011555,
+      "loss": 2.745,
+      "step": 17424
+    },
+    {
+      "epoch": 0.8885609726364334,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001230309154389708,
+      "loss": 2.7876,
+      "step": 17425
+    },
+    {
+      "epoch": 0.8886119660925388,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.001229609704297027,
+      "loss": 2.7562,
+      "step": 17426
+    },
+    {
+      "epoch": 0.8886629595486442,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0012289101742545684,
+      "loss": 2.7863,
+      "step": 17427
+    },
+    {
+      "epoch": 0.8887139530047495,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0012282105646938375,
+      "loss": 2.7887,
+      "step": 17428
+    },
+    {
+      "epoch": 0.8887649464608549,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0012275108760463884,
+      "loss": 2.775,
+      "step": 17429
+    },
+    {
+      "epoch": 0.8888159399169604,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0012268111087438242,
+      "loss": 2.7518,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8888669333730658,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0012261112632177974,
+      "loss": 2.7505,
+      "step": 17431
+    },
+    {
+      "epoch": 0.8889179268291711,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0012254113399000066,
+      "loss": 2.7913,
+      "step": 17432
+    },
+    {
+      "epoch": 0.8889689202852765,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0012247113392222005,
+      "loss": 2.7494,
+      "step": 17433
+    },
+    {
+      "epoch": 0.8890199137413819,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0012240112616161744,
+      "loss": 2.7742,
+      "step": 17434
+    },
+    {
+      "epoch": 0.8890709071974873,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0012233111075137713,
+      "loss": 2.7443,
+      "step": 17435
+    },
+    {
+      "epoch": 0.8891219006535926,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0012226108773468814,
+      "loss": 2.744,
+      "step": 17436
+    },
+    {
+      "epoch": 0.8891728941096981,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0012219105715474423,
+      "loss": 2.7575,
+      "step": 17437
+    },
+    {
+      "epoch": 0.8892238875658035,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0012212101905474373,
+      "loss": 2.7319,
+      "step": 17438
+    },
+    {
+      "epoch": 0.8892748810219089,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.001220509734778897,
+      "loss": 2.7418,
+      "step": 17439
+    },
+    {
+      "epoch": 0.8893258744780143,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.001219809204673897,
+      "loss": 2.7589,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8893768679341196,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0012191086006645604,
+      "loss": 2.7508,
+      "step": 17441
+    },
+    {
+      "epoch": 0.889427861390225,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0012184079231830542,
+      "loss": 2.7414,
+      "step": 17442
+    },
+    {
+      "epoch": 0.8894788548463304,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.001217707172661592,
+      "loss": 2.7527,
+      "step": 17443
+    },
+    {
+      "epoch": 0.8895298483024359,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0012170063495324315,
+      "loss": 2.7367,
+      "step": 17444
+    },
+    {
+      "epoch": 0.8895808417585412,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.001216305454227876,
+      "loss": 2.7567,
+      "step": 17445
+    },
+    {
+      "epoch": 0.8896318352146466,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0012156044871802728,
+      "loss": 2.794,
+      "step": 17446
+    },
+    {
+      "epoch": 0.889682828670752,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0012149034488220134,
+      "loss": 2.7427,
+      "step": 17447
+    },
+    {
+      "epoch": 0.8897338221268574,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0012142023395855333,
+      "loss": 2.7554,
+      "step": 17448
+    },
+    {
+      "epoch": 0.8897848155829627,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0012135011599033127,
+      "loss": 2.7604,
+      "step": 17449
+    },
+    {
+      "epoch": 0.8898358090390681,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.001212799910207874,
+      "loss": 2.7631,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8898868024951736,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0012120985909317829,
+      "loss": 2.7568,
+      "step": 17451
+    },
+    {
+      "epoch": 0.889937795951279,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0012113972025076488,
+      "loss": 2.799,
+      "step": 17452
+    },
+    {
+      "epoch": 0.8899887894073844,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0012106957453681235,
+      "loss": 2.7613,
+      "step": 17453
+    },
+    {
+      "epoch": 0.8900397828634897,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0012099942199459005,
+      "loss": 2.7575,
+      "step": 17454
+    },
+    {
+      "epoch": 0.8900907763195951,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0012092926266737162,
+      "loss": 2.7614,
+      "step": 17455
+    },
+    {
+      "epoch": 0.8901417697757005,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0012085909659843486,
+      "loss": 2.7577,
+      "step": 17456
+    },
+    {
+      "epoch": 0.8901927632318059,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0012078892383106173,
+      "loss": 2.7632,
+      "step": 17457
+    },
+    {
+      "epoch": 0.8902437566879113,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.001207187444085383,
+      "loss": 2.739,
+      "step": 17458
+    },
+    {
+      "epoch": 0.8902947501440167,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0012064855837415475,
+      "loss": 2.7408,
+      "step": 17459
+    },
+    {
+      "epoch": 0.8903457436001221,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.001205783657712054,
+      "loss": 2.7242,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8903967370562275,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0012050816664298854,
+      "loss": 2.7488,
+      "step": 17461
+    },
+    {
+      "epoch": 0.8904477305123328,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0012043796103280652,
+      "loss": 2.7723,
+      "step": 17462
+    },
+    {
+      "epoch": 0.8904987239684382,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0012036774898396571,
+      "loss": 2.7411,
+      "step": 17463
+    },
+    {
+      "epoch": 0.8905497174245436,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0012029753053977644,
+      "loss": 2.7916,
+      "step": 17464
+    },
+    {
+      "epoch": 0.8906007108806491,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0012022730574355292,
+      "loss": 2.7566,
+      "step": 17465
+    },
+    {
+      "epoch": 0.8906517043367544,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0012015707463861333,
+      "loss": 2.7535,
+      "step": 17466
+    },
+    {
+      "epoch": 0.8907026977928598,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0012008683726827984,
+      "loss": 2.7609,
+      "step": 17467
+    },
+    {
+      "epoch": 0.8907536912489652,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0012001659367587832,
+      "loss": 2.7435,
+      "step": 17468
+    },
+    {
+      "epoch": 0.8908046847050706,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0011994634390473853,
+      "loss": 2.7716,
+      "step": 17469
+    },
+    {
+      "epoch": 0.890855678161176,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0011987608799819407,
+      "loss": 2.7578,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8909066716172813,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0011980582599958236,
+      "loss": 2.7618,
+      "step": 17471
+    },
+    {
+      "epoch": 0.8909576650733868,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.001197355579522445,
+      "loss": 2.775,
+      "step": 17472
+    },
+    {
+      "epoch": 0.8910086585294922,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0011966528389952531,
+      "loss": 2.7451,
+      "step": 17473
+    },
+    {
+      "epoch": 0.8910596519855976,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.001195950038847734,
+      "loss": 2.7512,
+      "step": 17474
+    },
+    {
+      "epoch": 0.8911106454417029,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0011952471795134104,
+      "loss": 2.7509,
+      "step": 17475
+    },
+    {
+      "epoch": 0.8911616388978083,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0011945442614258407,
+      "loss": 2.7796,
+      "step": 17476
+    },
+    {
+      "epoch": 0.8912126323539137,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0011938412850186202,
+      "loss": 2.7748,
+      "step": 17477
+    },
+    {
+      "epoch": 0.8912636258100191,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0011931382507253804,
+      "loss": 2.747,
+      "step": 17478
+    },
+    {
+      "epoch": 0.8913146192661245,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.001192435158979788,
+      "loss": 2.7427,
+      "step": 17479
+    },
+    {
+      "epoch": 0.8913656127222299,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001191732010215545,
+      "loss": 2.7367,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8914166061783353,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0011910288048663893,
+      "loss": 2.7635,
+      "step": 17481
+    },
+    {
+      "epoch": 0.8914675996344407,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0011903255433660933,
+      "loss": 2.7445,
+      "step": 17482
+    },
+    {
+      "epoch": 0.891518593090546,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0011896222261484639,
+      "loss": 2.7504,
+      "step": 17483
+    },
+    {
+      "epoch": 0.8915695865466514,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0011889188536473424,
+      "loss": 2.7412,
+      "step": 17484
+    },
+    {
+      "epoch": 0.8916205800027568,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0011882154262966048,
+      "loss": 2.7355,
+      "step": 17485
+    },
+    {
+      "epoch": 0.8916715734588623,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0011875119445301596,
+      "loss": 2.7627,
+      "step": 17486
+    },
+    {
+      "epoch": 0.8917225669149677,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011868084087819503,
+      "loss": 2.7779,
+      "step": 17487
+    },
+    {
+      "epoch": 0.891773560371073,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0011861048194859527,
+      "loss": 2.7743,
+      "step": 17488
+    },
+    {
+      "epoch": 0.8918245538271784,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0011854011770761764,
+      "loss": 2.752,
+      "step": 17489
+    },
+    {
+      "epoch": 0.8918755472832838,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0011846974819866632,
+      "loss": 2.7477,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8919265407393892,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0011839937346514874,
+      "loss": 2.7649,
+      "step": 17491
+    },
+    {
+      "epoch": 0.8919775341954945,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0011832899355047563,
+      "loss": 2.7625,
+      "step": 17492
+    },
+    {
+      "epoch": 0.8920285276516,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0011825860849806078,
+      "loss": 2.7481,
+      "step": 17493
+    },
+    {
+      "epoch": 0.8920795211077054,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011818821835132133,
+      "loss": 2.7659,
+      "step": 17494
+    },
+    {
+      "epoch": 0.8921305145638108,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0011811782315367732,
+      "loss": 2.75,
+      "step": 17495
+    },
+    {
+      "epoch": 0.8921815080199161,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0011804742294855217,
+      "loss": 2.7333,
+      "step": 17496
+    },
+    {
+      "epoch": 0.8922325014760215,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011797701777937221,
+      "loss": 2.7333,
+      "step": 17497
+    },
+    {
+      "epoch": 0.8922834949321269,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0011790660768956692,
+      "loss": 2.778,
+      "step": 17498
+    },
+    {
+      "epoch": 0.8923344883882323,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.001178361927225687,
+      "loss": 2.7314,
+      "step": 17499
+    },
+    {
+      "epoch": 0.8923854818443377,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0011776577292181311,
+      "loss": 2.7759,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8924364753004431,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.001176953483307386,
+      "loss": 2.762,
+      "step": 17501
+    },
+    {
+      "epoch": 0.8924874687565485,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0011762491899278663,
+      "loss": 2.7528,
+      "step": 17502
+    },
+    {
+      "epoch": 0.8925384622126539,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0011755448495140148,
+      "loss": 2.7652,
+      "step": 17503
+    },
+    {
+      "epoch": 0.8925894556687592,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0011748404625003051,
+      "loss": 2.7409,
+      "step": 17504
+    },
+    {
+      "epoch": 0.8926404491248646,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0011741360293212372,
+      "loss": 2.7672,
+      "step": 17505
+    },
+    {
+      "epoch": 0.89269144258097,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0011734315504113419,
+      "loss": 2.7584,
+      "step": 17506
+    },
+    {
+      "epoch": 0.8927424360370755,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.001172727026205177,
+      "loss": 2.7473,
+      "step": 17507
+    },
+    {
+      "epoch": 0.8927934294931809,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0011720224571373282,
+      "loss": 2.7631,
+      "step": 17508
+    },
+    {
+      "epoch": 0.8928444229492862,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0011713178436424088,
+      "loss": 2.7418,
+      "step": 17509
+    },
+    {
+      "epoch": 0.8928954164053916,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0011706131861550606,
+      "loss": 2.7523,
+      "step": 17510
+    },
+    {
+      "epoch": 0.892946409861497,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0011699084851099512,
+      "loss": 2.7403,
+      "step": 17511
+    },
+    {
+      "epoch": 0.8929974033176024,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.001169203740941776,
+      "loss": 2.7547,
+      "step": 17512
+    },
+    {
+      "epoch": 0.8930483967737077,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011684989540852557,
+      "loss": 2.7682,
+      "step": 17513
+    },
+    {
+      "epoch": 0.8930993902298132,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0011677941249751396,
+      "loss": 2.7459,
+      "step": 17514
+    },
+    {
+      "epoch": 0.8931503836859186,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011670892540462008,
+      "loss": 2.7666,
+      "step": 17515
+    },
+    {
+      "epoch": 0.893201377142024,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.001166384341733239,
+      "loss": 2.7719,
+      "step": 17516
+    },
+    {
+      "epoch": 0.8932523705981293,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0011656793884710802,
+      "loss": 2.7363,
+      "step": 17517
+    },
+    {
+      "epoch": 0.8933033640542347,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0011649743946945748,
+      "loss": 2.7536,
+      "step": 17518
+    },
+    {
+      "epoch": 0.8933543575103401,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0011642693608385978,
+      "loss": 2.7731,
+      "step": 17519
+    },
+    {
+      "epoch": 0.8934053509664455,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00116356428733805,
+      "loss": 2.7285,
+      "step": 17520
+    },
+    {
+      "epoch": 0.893456344422551,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001162859174627856,
+      "loss": 2.775,
+      "step": 17521
+    },
+    {
+      "epoch": 0.8935073378786563,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0011621540231429655,
+      "loss": 2.7705,
+      "step": 17522
+    },
+    {
+      "epoch": 0.8935583313347617,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0011614488333183501,
+      "loss": 2.7584,
+      "step": 17523
+    },
+    {
+      "epoch": 0.8936093247908671,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.001160743605589007,
+      "loss": 2.7658,
+      "step": 17524
+    },
+    {
+      "epoch": 0.8936603182469725,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0011600383403899562,
+      "loss": 2.7698,
+      "step": 17525
+    },
+    {
+      "epoch": 0.8937113117030778,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0011593330381562401,
+      "loss": 2.7177,
+      "step": 17526
+    },
+    {
+      "epoch": 0.8937623051591832,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0011586276993229253,
+      "loss": 2.7483,
+      "step": 17527
+    },
+    {
+      "epoch": 0.8938132986152887,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0011579223243250998,
+      "loss": 2.74,
+      "step": 17528
+    },
+    {
+      "epoch": 0.8938642920713941,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0011572169135978744,
+      "loss": 2.7658,
+      "step": 17529
+    },
+    {
+      "epoch": 0.8939152855274994,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0011565114675763822,
+      "loss": 2.7288,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8939662789836048,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0011558059866957774,
+      "loss": 2.7597,
+      "step": 17531
+    },
+    {
+      "epoch": 0.8940172724397102,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0011551004713912358,
+      "loss": 2.7829,
+      "step": 17532
+    },
+    {
+      "epoch": 0.8940682658958156,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0011543949220979553,
+      "loss": 2.7824,
+      "step": 17533
+    },
+    {
+      "epoch": 0.8941192593519209,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.001153689339251154,
+      "loss": 2.7506,
+      "step": 17534
+    },
+    {
+      "epoch": 0.8941702528080264,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0011529837232860701,
+      "loss": 2.7606,
+      "step": 17535
+    },
+    {
+      "epoch": 0.8942212462641318,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0011522780746379643,
+      "loss": 2.7556,
+      "step": 17536
+    },
+    {
+      "epoch": 0.8942722397202372,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0011515723937421149,
+      "loss": 2.7308,
+      "step": 17537
+    },
+    {
+      "epoch": 0.8943232331763425,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0011508666810338222,
+      "loss": 2.761,
+      "step": 17538
+    },
+    {
+      "epoch": 0.8943742266324479,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0011501609369484046,
+      "loss": 2.7691,
+      "step": 17539
+    },
+    {
+      "epoch": 0.8944252200885533,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0011494551619212012,
+      "loss": 2.7459,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8944762135446587,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001148749356387569,
+      "loss": 2.7676,
+      "step": 17541
+    },
+    {
+      "epoch": 0.8945272070007642,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0011480435207828845,
+      "loss": 2.7516,
+      "step": 17542
+    },
+    {
+      "epoch": 0.8945782004568695,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0011473376555425426,
+      "loss": 2.7577,
+      "step": 17543
+    },
+    {
+      "epoch": 0.8946291939129749,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0011466317611019564,
+      "loss": 2.7646,
+      "step": 17544
+    },
+    {
+      "epoch": 0.8946801873690803,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0011459258378965565,
+      "loss": 2.7296,
+      "step": 17545
+    },
+    {
+      "epoch": 0.8947311808251857,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0011452198863617926,
+      "loss": 2.7688,
+      "step": 17546
+    },
+    {
+      "epoch": 0.894782174281291,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.001144513906933131,
+      "loss": 2.7193,
+      "step": 17547
+    },
+    {
+      "epoch": 0.8948331677373964,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.001143807900046055,
+      "loss": 2.7582,
+      "step": 17548
+    },
+    {
+      "epoch": 0.8948841611935019,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0011431018661360649,
+      "loss": 2.763,
+      "step": 17549
+    },
+    {
+      "epoch": 0.8949351546496073,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0011423958056386782,
+      "loss": 2.7468,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8949861481057126,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0011416897189894292,
+      "loss": 2.7492,
+      "step": 17551
+    },
+    {
+      "epoch": 0.895037141561818,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0011409836066238663,
+      "loss": 2.7837,
+      "step": 17552
+    },
+    {
+      "epoch": 0.8950881350179234,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0011402774689775554,
+      "loss": 2.7526,
+      "step": 17553
+    },
+    {
+      "epoch": 0.8951391284740288,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0011395713064860785,
+      "loss": 2.7357,
+      "step": 17554
+    },
+    {
+      "epoch": 0.8951901219301341,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0011388651195850313,
+      "loss": 2.7575,
+      "step": 17555
+    },
+    {
+      "epoch": 0.8952411153862396,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0011381589087100259,
+      "loss": 2.7319,
+      "step": 17556
+    },
+    {
+      "epoch": 0.895292108842345,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0011374526742966879,
+      "loss": 2.7526,
+      "step": 17557
+    },
+    {
+      "epoch": 0.8953431022984504,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0011367464167806591,
+      "loss": 2.7332,
+      "step": 17558
+    },
+    {
+      "epoch": 0.8953940957545558,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0011360401365975942,
+      "loss": 2.7577,
+      "step": 17559
+    },
+    {
+      "epoch": 0.8954450892106611,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0011353338341831619,
+      "loss": 2.7539,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8954960826667665,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0011346275099730454,
+      "loss": 2.72,
+      "step": 17561
+    },
+    {
+      "epoch": 0.8955470761228719,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0011339211644029413,
+      "loss": 2.7677,
+      "step": 17562
+    },
+    {
+      "epoch": 0.8955980695789774,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0011332147979085582,
+      "loss": 2.7411,
+      "step": 17563
+    },
+    {
+      "epoch": 0.8956490630350827,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0011325084109256194,
+      "loss": 2.7414,
+      "step": 17564
+    },
+    {
+      "epoch": 0.8957000564911881,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.001131802003889859,
+      "loss": 2.7377,
+      "step": 17565
+    },
+    {
+      "epoch": 0.8957510499472935,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0011310955772370253,
+      "loss": 2.7396,
+      "step": 17566
+    },
+    {
+      "epoch": 0.8958020434033989,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0011303891314028767,
+      "loss": 2.7557,
+      "step": 17567
+    },
+    {
+      "epoch": 0.8958530368595042,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0011296826668231851,
+      "loss": 2.747,
+      "step": 17568
+    },
+    {
+      "epoch": 0.8959040303156096,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0011289761839337335,
+      "loss": 2.7287,
+      "step": 17569
+    },
+    {
+      "epoch": 0.8959550237717151,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0011282696831703153,
+      "loss": 2.7606,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8960060172278205,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0011275631649687366,
+      "loss": 2.7492,
+      "step": 17571
+    },
+    {
+      "epoch": 0.8960570106839258,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0011268566297648129,
+      "loss": 2.7663,
+      "step": 17572
+    },
+    {
+      "epoch": 0.8961080041400312,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0011261500779943705,
+      "loss": 2.7725,
+      "step": 17573
+    },
+    {
+      "epoch": 0.8961589975961366,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0011254435100932461,
+      "loss": 2.7506,
+      "step": 17574
+    },
+    {
+      "epoch": 0.896209991052242,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.001124736926497286,
+      "loss": 2.7659,
+      "step": 17575
+    },
+    {
+      "epoch": 0.8962609845083473,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0011240303276423473,
+      "loss": 2.747,
+      "step": 17576
+    },
+    {
+      "epoch": 0.8963119779644528,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0011233237139642949,
+      "loss": 2.7449,
+      "step": 17577
+    },
+    {
+      "epoch": 0.8963629714205582,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011226170858990039,
+      "loss": 2.7454,
+      "step": 17578
+    },
+    {
+      "epoch": 0.8964139648766636,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.001121910443882358,
+      "loss": 2.7403,
+      "step": 17579
+    },
+    {
+      "epoch": 0.896464958332769,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0011212037883502493,
+      "loss": 2.7431,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8965159517888743,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0011204971197385785,
+      "loss": 2.7236,
+      "step": 17581
+    },
+    {
+      "epoch": 0.8965669452449797,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0011197904384832548,
+      "loss": 2.7486,
+      "step": 17582
+    },
+    {
+      "epoch": 0.8966179387010851,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0011190837450201939,
+      "loss": 2.7533,
+      "step": 17583
+    },
+    {
+      "epoch": 0.8966689321571906,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0011183770397853205,
+      "loss": 2.7657,
+      "step": 17584
+    },
+    {
+      "epoch": 0.8967199256132959,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0011176703232145654,
+      "loss": 2.7627,
+      "step": 17585
+    },
+    {
+      "epoch": 0.8967709190694013,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0011169635957438675,
+      "loss": 2.7372,
+      "step": 17586
+    },
+    {
+      "epoch": 0.8968219125255067,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0011162568578091713,
+      "loss": 2.7378,
+      "step": 17587
+    },
+    {
+      "epoch": 0.8968729059816121,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0011155501098464283,
+      "loss": 2.7306,
+      "step": 17588
+    },
+    {
+      "epoch": 0.8969238994377174,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0011148433522915962,
+      "loss": 2.7582,
+      "step": 17589
+    },
+    {
+      "epoch": 0.8969748928938228,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0011141365855806386,
+      "loss": 2.7434,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8970258863499283,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0011134298101495251,
+      "loss": 2.7415,
+      "step": 17591
+    },
+    {
+      "epoch": 0.8970768798060337,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0011127230264342297,
+      "loss": 2.7413,
+      "step": 17592
+    },
+    {
+      "epoch": 0.897127873262139,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0011120162348707322,
+      "loss": 2.7814,
+      "step": 17593
+    },
+    {
+      "epoch": 0.8971788667182444,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0011113094358950176,
+      "loss": 2.7849,
+      "step": 17594
+    },
+    {
+      "epoch": 0.8972298601743498,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0011106026299430743,
+      "loss": 2.7444,
+      "step": 17595
+    },
+    {
+      "epoch": 0.8972808536304552,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0011098958174508963,
+      "loss": 2.7469,
+      "step": 17596
+    },
+    {
+      "epoch": 0.8973318470865606,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0011091889988544812,
+      "loss": 2.7486,
+      "step": 17597
+    },
+    {
+      "epoch": 0.897382840542666,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0011084821745898298,
+      "loss": 2.7477,
+      "step": 17598
+    },
+    {
+      "epoch": 0.8974338339987714,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0011077753450929467,
+      "loss": 2.762,
+      "step": 17599
+    },
+    {
+      "epoch": 0.8974848274548768,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0011070685107998403,
+      "loss": 2.7619,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8975358209109822,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0011063616721465213,
+      "loss": 2.7438,
+      "step": 17601
+    },
+    {
+      "epoch": 0.8975868143670875,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0011056548295690033,
+      "loss": 2.7592,
+      "step": 17602
+    },
+    {
+      "epoch": 0.8976378078231929,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0011049479835033018,
+      "loss": 2.7368,
+      "step": 17603
+    },
+    {
+      "epoch": 0.8976888012792983,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0011042411343854359,
+      "loss": 2.758,
+      "step": 17604
+    },
+    {
+      "epoch": 0.8977397947354038,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0011035342826514255,
+      "loss": 2.7318,
+      "step": 17605
+    },
+    {
+      "epoch": 0.8977907881915091,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0011028274287372918,
+      "loss": 2.7478,
+      "step": 17606
+    },
+    {
+      "epoch": 0.8978417816476145,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0011021205730790578,
+      "loss": 2.7587,
+      "step": 17607
+    },
+    {
+      "epoch": 0.8978927751037199,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0011014137161127477,
+      "loss": 2.7478,
+      "step": 17608
+    },
+    {
+      "epoch": 0.8979437685598253,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0011007068582743868,
+      "loss": 2.7605,
+      "step": 17609
+    },
+    {
+      "epoch": 0.8979947620159306,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0011,
+      "loss": 2.7271,
+      "step": 17610
+    },
+    {
+      "epoch": 0.898045755472036,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0010992931417256133,
+      "loss": 2.7273,
+      "step": 17611
+    },
+    {
+      "epoch": 0.8980967489281415,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0010985862838872524,
+      "loss": 2.736,
+      "step": 17612
+    },
+    {
+      "epoch": 0.8981477423842469,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0010978794269209423,
+      "loss": 2.738,
+      "step": 17613
+    },
+    {
+      "epoch": 0.8981987358403523,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0010971725712627086,
+      "loss": 2.7785,
+      "step": 17614
+    },
+    {
+      "epoch": 0.8982497292964576,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0010964657173485749,
+      "loss": 2.7563,
+      "step": 17615
+    },
+    {
+      "epoch": 0.898300722752563,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001095758865614564,
+      "loss": 2.7861,
+      "step": 17616
+    },
+    {
+      "epoch": 0.8983517162086684,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.001095052016496698,
+      "loss": 2.7488,
+      "step": 17617
+    },
+    {
+      "epoch": 0.8984027096647738,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.001094345170430997,
+      "loss": 2.7487,
+      "step": 17618
+    },
+    {
+      "epoch": 0.8984537031208792,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.001093638327853479,
+      "loss": 2.7425,
+      "step": 17619
+    },
+    {
+      "epoch": 0.8985046965769846,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0010929314892001603,
+      "loss": 2.7384,
+      "step": 17620
+    },
+    {
+      "epoch": 0.89855569003309,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0010922246549070534,
+      "loss": 2.7588,
+      "step": 17621
+    },
+    {
+      "epoch": 0.8986066834891954,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0010915178254101705,
+      "loss": 2.7418,
+      "step": 17622
+    },
+    {
+      "epoch": 0.8986576769453007,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001090811001145519,
+      "loss": 2.7044,
+      "step": 17623
+    },
+    {
+      "epoch": 0.8987086704014061,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0010901041825491038,
+      "loss": 2.7319,
+      "step": 17624
+    },
+    {
+      "epoch": 0.8987596638575115,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0010893973700569258,
+      "loss": 2.7646,
+      "step": 17625
+    },
+    {
+      "epoch": 0.898810657313617,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0010886905641049828,
+      "loss": 2.774,
+      "step": 17626
+    },
+    {
+      "epoch": 0.8988616507697224,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0010879837651292682,
+      "loss": 2.7636,
+      "step": 17627
+    },
+    {
+      "epoch": 0.8989126442258277,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0010872769735657708,
+      "loss": 2.7498,
+      "step": 17628
+    },
+    {
+      "epoch": 0.8989636376819331,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.001086570189850475,
+      "loss": 2.7512,
+      "step": 17629
+    },
+    {
+      "epoch": 0.8990146311380385,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0010858634144193615,
+      "loss": 2.766,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8990656245941439,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0010851566477084042,
+      "loss": 2.7469,
+      "step": 17631
+    },
+    {
+      "epoch": 0.8991166180502492,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0010844498901535723,
+      "loss": 2.7241,
+      "step": 17632
+    },
+    {
+      "epoch": 0.8991676115063547,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0010837431421908288,
+      "loss": 2.7392,
+      "step": 17633
+    },
+    {
+      "epoch": 0.8992186049624601,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0010830364042561328,
+      "loss": 2.7695,
+      "step": 17634
+    },
+    {
+      "epoch": 0.8992695984185655,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0010823296767854347,
+      "loss": 2.7097,
+      "step": 17635
+    },
+    {
+      "epoch": 0.8993205918746708,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0010816229602146799,
+      "loss": 2.735,
+      "step": 17636
+    },
+    {
+      "epoch": 0.8993715853307762,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0010809162549798063,
+      "loss": 2.7309,
+      "step": 17637
+    },
+    {
+      "epoch": 0.8994225787868816,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0010802095615167456,
+      "loss": 2.739,
+      "step": 17638
+    },
+    {
+      "epoch": 0.899473572242987,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0010795028802614217,
+      "loss": 2.7616,
+      "step": 17639
+    },
+    {
+      "epoch": 0.8995245656990924,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001078796211649751,
+      "loss": 2.7373,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8995755591551978,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0010780895561176423,
+      "loss": 2.7597,
+      "step": 17641
+    },
+    {
+      "epoch": 0.8996265526113032,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0010773829141009963,
+      "loss": 2.732,
+      "step": 17642
+    },
+    {
+      "epoch": 0.8996775460674086,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0010766762860357053,
+      "loss": 2.7603,
+      "step": 17643
+    },
+    {
+      "epoch": 0.899728539523514,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.001075969672357653,
+      "loss": 2.7447,
+      "step": 17644
+    },
+    {
+      "epoch": 0.8997795329796193,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.001075263073502714,
+      "loss": 2.7629,
+      "step": 17645
+    },
+    {
+      "epoch": 0.8998305264357247,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0010745564899067542,
+      "loss": 2.7752,
+      "step": 17646
+    },
+    {
+      "epoch": 0.8998815198918302,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0010738499220056296,
+      "loss": 2.7675,
+      "step": 17647
+    },
+    {
+      "epoch": 0.8999325133479356,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0010731433702351873,
+      "loss": 2.7354,
+      "step": 17648
+    },
+    {
+      "epoch": 0.8999835068040409,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0010724368350312635,
+      "loss": 2.7296,
+      "step": 17649
+    },
+    {
+      "epoch": 0.9000345002601463,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0010717303168296846,
+      "loss": 2.7691,
+      "step": 17650
+    },
+    {
+      "epoch": 0.9000854937162517,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0010710238160662666,
+      "loss": 2.7332,
+      "step": 17651
+    },
+    {
+      "epoch": 0.9001364871723571,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001070317333176815,
+      "loss": 2.723,
+      "step": 17652
+    },
+    {
+      "epoch": 0.9001874806284624,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0010696108685971234,
+      "loss": 2.7578,
+      "step": 17653
+    },
+    {
+      "epoch": 0.9002384740845679,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0010689044227629749,
+      "loss": 2.7159,
+      "step": 17654
+    },
+    {
+      "epoch": 0.9002894675406733,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0010681979961101409,
+      "loss": 2.736,
+      "step": 17655
+    },
+    {
+      "epoch": 0.9003404609967787,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0010674915890743811,
+      "loss": 2.7614,
+      "step": 17656
+    },
+    {
+      "epoch": 0.900391454452884,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0010667852020914417,
+      "loss": 2.7474,
+      "step": 17657
+    },
+    {
+      "epoch": 0.9004424479089894,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.001066078835597059,
+      "loss": 2.7244,
+      "step": 17658
+    },
+    {
+      "epoch": 0.9004934413650948,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0010653724900269545,
+      "loss": 2.7658,
+      "step": 17659
+    },
+    {
+      "epoch": 0.9005444348212002,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0010646661658168387,
+      "loss": 2.7472,
+      "step": 17660
+    },
+    {
+      "epoch": 0.9005954282773057,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.001063959863402406,
+      "loss": 2.7632,
+      "step": 17661
+    },
+    {
+      "epoch": 0.900646421733411,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.001063253583219341,
+      "loss": 2.7452,
+      "step": 17662
+    },
+    {
+      "epoch": 0.9006974151895164,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0010625473257033123,
+      "loss": 2.7533,
+      "step": 17663
+    },
+    {
+      "epoch": 0.9007484086456218,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0010618410912899745,
+      "loss": 2.755,
+      "step": 17664
+    },
+    {
+      "epoch": 0.9007994021017272,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0010611348804149688,
+      "loss": 2.7454,
+      "step": 17665
+    },
+    {
+      "epoch": 0.9008503955578325,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0010604286935139218,
+      "loss": 2.729,
+      "step": 17666
+    },
+    {
+      "epoch": 0.9009013890139379,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0010597225310224447,
+      "loss": 2.7191,
+      "step": 17667
+    },
+    {
+      "epoch": 0.9009523824700434,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0010590163933761343,
+      "loss": 2.7468,
+      "step": 17668
+    },
+    {
+      "epoch": 0.9010033759261488,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.001058310281010571,
+      "loss": 2.7407,
+      "step": 17669
+    },
+    {
+      "epoch": 0.9010543693822541,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0010576041943613217,
+      "loss": 2.7758,
+      "step": 17670
+    },
+    {
+      "epoch": 0.9011053628383595,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0010568981338639353,
+      "loss": 2.7676,
+      "step": 17671
+    },
+    {
+      "epoch": 0.9011563562944649,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0010561920999539454,
+      "loss": 2.7546,
+      "step": 17672
+    },
+    {
+      "epoch": 0.9012073497505703,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0010554860930668693,
+      "loss": 2.741,
+      "step": 17673
+    },
+    {
+      "epoch": 0.9012583432066756,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0010547801136382075,
+      "loss": 2.7498,
+      "step": 17674
+    },
+    {
+      "epoch": 0.9013093366627811,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0010540741621034438,
+      "loss": 2.7538,
+      "step": 17675
+    },
+    {
+      "epoch": 0.9013603301188865,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0010533682388980442,
+      "loss": 2.7688,
+      "step": 17676
+    },
+    {
+      "epoch": 0.9014113235749919,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0010526623444574578,
+      "loss": 2.7517,
+      "step": 17677
+    },
+    {
+      "epoch": 0.9014623170310972,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0010519564792171156,
+      "loss": 2.7238,
+      "step": 17678
+    },
+    {
+      "epoch": 0.9015133104872026,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0010512506436124312,
+      "loss": 2.7475,
+      "step": 17679
+    },
+    {
+      "epoch": 0.901564303943308,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.001050544838078799,
+      "loss": 2.7576,
+      "step": 17680
+    },
+    {
+      "epoch": 0.9016152973994134,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0010498390630515953,
+      "loss": 2.7332,
+      "step": 17681
+    },
+    {
+      "epoch": 0.9016662908555189,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001049133318966178,
+      "loss": 2.7435,
+      "step": 17682
+    },
+    {
+      "epoch": 0.9017172843116242,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0010484276062578853,
+      "loss": 2.7421,
+      "step": 17683
+    },
+    {
+      "epoch": 0.9017682777677296,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.001047721925362036,
+      "loss": 2.7449,
+      "step": 17684
+    },
+    {
+      "epoch": 0.901819271223835,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0010470162767139298,
+      "loss": 2.745,
+      "step": 17685
+    },
+    {
+      "epoch": 0.9018702646799404,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0010463106607488464,
+      "loss": 2.7432,
+      "step": 17686
+    },
+    {
+      "epoch": 0.9019212581360457,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0010456050779020449,
+      "loss": 2.7352,
+      "step": 17687
+    },
+    {
+      "epoch": 0.9019722515921511,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0010448995286087644,
+      "loss": 2.7124,
+      "step": 17688
+    },
+    {
+      "epoch": 0.9020232450482566,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001044194013304223,
+      "loss": 2.7625,
+      "step": 17689
+    },
+    {
+      "epoch": 0.902074238504362,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0010434885324236182,
+      "loss": 2.7365,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9021252319604673,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0010427830864021257,
+      "loss": 2.7448,
+      "step": 17691
+    },
+    {
+      "epoch": 0.9021762254165727,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0010420776756749003,
+      "loss": 2.7328,
+      "step": 17692
+    },
+    {
+      "epoch": 0.9022272188726781,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0010413723006770749,
+      "loss": 2.7465,
+      "step": 17693
+    },
+    {
+      "epoch": 0.9022782123287835,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0010406669618437598,
+      "loss": 2.7392,
+      "step": 17694
+    },
+    {
+      "epoch": 0.9023292057848888,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0010399616596100444,
+      "loss": 2.7514,
+      "step": 17695
+    },
+    {
+      "epoch": 0.9023801992409943,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0010392563944109934,
+      "loss": 2.7478,
+      "step": 17696
+    },
+    {
+      "epoch": 0.9024311926970997,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00103855116668165,
+      "loss": 2.7341,
+      "step": 17697
+    },
+    {
+      "epoch": 0.9024821861532051,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0010378459768570348,
+      "loss": 2.7427,
+      "step": 17698
+    },
+    {
+      "epoch": 0.9025331796093105,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0010371408253721439,
+      "loss": 2.7453,
+      "step": 17699
+    },
+    {
+      "epoch": 0.9025841730654158,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0010364357126619501,
+      "loss": 2.7102,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9026351665215212,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.001035730639161402,
+      "loss": 2.7302,
+      "step": 17701
+    },
+    {
+      "epoch": 0.9026861599776266,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0010350256053054255,
+      "loss": 2.7433,
+      "step": 17702
+    },
+    {
+      "epoch": 0.9027371534337321,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0010343206115289202,
+      "loss": 2.7237,
+      "step": 17703
+    },
+    {
+      "epoch": 0.9027881468898374,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0010336156582667613,
+      "loss": 2.7628,
+      "step": 17704
+    },
+    {
+      "epoch": 0.9028391403459428,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0010329107459537993,
+      "loss": 2.7582,
+      "step": 17705
+    },
+    {
+      "epoch": 0.9028901338020482,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0010322058750248605,
+      "loss": 2.7328,
+      "step": 17706
+    },
+    {
+      "epoch": 0.9029411272581536,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0010315010459147442,
+      "loss": 2.7526,
+      "step": 17707
+    },
+    {
+      "epoch": 0.9029921207142589,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0010307962590582246,
+      "loss": 2.7576,
+      "step": 17708
+    },
+    {
+      "epoch": 0.9030431141703643,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.001030091514890049,
+      "loss": 2.7699,
+      "step": 17709
+    },
+    {
+      "epoch": 0.9030941076264698,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0010293868138449398,
+      "loss": 2.6925,
+      "step": 17710
+    },
+    {
+      "epoch": 0.9031451010825752,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0010286821563575915,
+      "loss": 2.7364,
+      "step": 17711
+    },
+    {
+      "epoch": 0.9031960945386805,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0010279775428626724,
+      "loss": 2.7537,
+      "step": 17712
+    },
+    {
+      "epoch": 0.9032470879947859,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0010272729737948233,
+      "loss": 2.752,
+      "step": 17713
+    },
+    {
+      "epoch": 0.9032980814508913,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0010265684495886582,
+      "loss": 2.7454,
+      "step": 17714
+    },
+    {
+      "epoch": 0.9033490749069967,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0010258639706787632,
+      "loss": 2.7254,
+      "step": 17715
+    },
+    {
+      "epoch": 0.903400068363102,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0010251595374996955,
+      "loss": 2.762,
+      "step": 17716
+    },
+    {
+      "epoch": 0.9034510618192075,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.001024455150485985,
+      "loss": 2.7502,
+      "step": 17717
+    },
+    {
+      "epoch": 0.9035020552753129,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001023750810072134,
+      "loss": 2.7495,
+      "step": 17718
+    },
+    {
+      "epoch": 0.9035530487314183,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0010230465166926142,
+      "loss": 2.7489,
+      "step": 17719
+    },
+    {
+      "epoch": 0.9036040421875237,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001022342270781869,
+      "loss": 2.7767,
+      "step": 17720
+    },
+    {
+      "epoch": 0.903655035643629,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0010216380727743131,
+      "loss": 2.7397,
+      "step": 17721
+    },
+    {
+      "epoch": 0.9037060290997344,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0010209339231043314,
+      "loss": 2.7407,
+      "step": 17722
+    },
+    {
+      "epoch": 0.9037570225558398,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.001020229822206278,
+      "loss": 2.7311,
+      "step": 17723
+    },
+    {
+      "epoch": 0.9038080160119453,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0010195257705144785,
+      "loss": 2.7372,
+      "step": 17724
+    },
+    {
+      "epoch": 0.9038590094680506,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.001018821768463227,
+      "loss": 2.7568,
+      "step": 17725
+    },
+    {
+      "epoch": 0.903910002924156,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.001018117816486787,
+      "loss": 2.7562,
+      "step": 17726
+    },
+    {
+      "epoch": 0.9039609963802614,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0010174139150193923,
+      "loss": 2.7295,
+      "step": 17727
+    },
+    {
+      "epoch": 0.9040119898363668,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.001016710064495244,
+      "loss": 2.7596,
+      "step": 17728
+    },
+    {
+      "epoch": 0.9040629832924721,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0010160062653485128,
+      "loss": 2.7208,
+      "step": 17729
+    },
+    {
+      "epoch": 0.9041139767485775,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0010153025180133371,
+      "loss": 2.7482,
+      "step": 17730
+    },
+    {
+      "epoch": 0.904164970204683,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.001014598822923824,
+      "loss": 2.7262,
+      "step": 17731
+    },
+    {
+      "epoch": 0.9042159636607884,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0010138951805140476,
+      "loss": 2.7541,
+      "step": 17732
+    },
+    {
+      "epoch": 0.9042669571168938,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0010131915912180499,
+      "loss": 2.7572,
+      "step": 17733
+    },
+    {
+      "epoch": 0.9043179505729991,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0010124880554698405,
+      "loss": 2.7627,
+      "step": 17734
+    },
+    {
+      "epoch": 0.9043689440291045,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0010117845737033955,
+      "loss": 2.752,
+      "step": 17735
+    },
+    {
+      "epoch": 0.9044199374852099,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0010110811463526577,
+      "loss": 2.7427,
+      "step": 17736
+    },
+    {
+      "epoch": 0.9044709309413153,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0010103777738515363,
+      "loss": 2.7755,
+      "step": 17737
+    },
+    {
+      "epoch": 0.9045219243974207,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0010096744566339064,
+      "loss": 2.7945,
+      "step": 17738
+    },
+    {
+      "epoch": 0.9045729178535261,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0010089711951336107,
+      "loss": 2.7558,
+      "step": 17739
+    },
+    {
+      "epoch": 0.9046239113096315,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001008267989784455,
+      "loss": 2.7529,
+      "step": 17740
+    },
+    {
+      "epoch": 0.9046749047657369,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0010075648410202125,
+      "loss": 2.742,
+      "step": 17741
+    },
+    {
+      "epoch": 0.9047258982218422,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00100686174927462,
+      "loss": 2.7157,
+      "step": 17742
+    },
+    {
+      "epoch": 0.9047768916779476,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00100615871498138,
+      "loss": 2.7564,
+      "step": 17743
+    },
+    {
+      "epoch": 0.904827885134053,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0010054557385741599,
+      "loss": 2.7656,
+      "step": 17744
+    },
+    {
+      "epoch": 0.9048788785901585,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0010047528204865902,
+      "loss": 2.7743,
+      "step": 17745
+    },
+    {
+      "epoch": 0.9049298720462639,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0010040499611522662,
+      "loss": 2.7222,
+      "step": 17746
+    },
+    {
+      "epoch": 0.9049808655023692,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001003347161004747,
+      "loss": 2.7247,
+      "step": 17747
+    },
+    {
+      "epoch": 0.9050318589584746,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0010026444204775556,
+      "loss": 2.7559,
+      "step": 17748
+    },
+    {
+      "epoch": 0.90508285241458,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0010019417400041768,
+      "loss": 2.7291,
+      "step": 17749
+    },
+    {
+      "epoch": 0.9051338458706853,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0010012391200180592,
+      "loss": 2.7299,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9051848393267907,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0010005365609526149,
+      "loss": 2.7407,
+      "step": 17751
+    },
+    {
+      "epoch": 0.9052358327828962,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0009998340632412172,
+      "loss": 2.7616,
+      "step": 17752
+    },
+    {
+      "epoch": 0.9052868262390016,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0009991316273172018,
+      "loss": 2.7544,
+      "step": 17753
+    },
+    {
+      "epoch": 0.905337819695107,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0009984292536138666,
+      "loss": 2.767,
+      "step": 17754
+    },
+    {
+      "epoch": 0.9053888131512123,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0009977269425644711,
+      "loss": 2.7485,
+      "step": 17755
+    },
+    {
+      "epoch": 0.9054398066073177,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.000997024694602236,
+      "loss": 2.7395,
+      "step": 17756
+    },
+    {
+      "epoch": 0.9054908000634231,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000996322510160343,
+      "loss": 2.754,
+      "step": 17757
+    },
+    {
+      "epoch": 0.9055417935195285,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.000995620389671935,
+      "loss": 2.7234,
+      "step": 17758
+    },
+    {
+      "epoch": 0.905592786975634,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0009949183335701147,
+      "loss": 2.7412,
+      "step": 17759
+    },
+    {
+      "epoch": 0.9056437804317393,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0009942163422879463,
+      "loss": 2.7278,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9056947738878447,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0009935144162584526,
+      "loss": 2.7635,
+      "step": 17761
+    },
+    {
+      "epoch": 0.9057457673439501,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0009928125559146173,
+      "loss": 2.7766,
+      "step": 17762
+    },
+    {
+      "epoch": 0.9057967608000554,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000992110761689383,
+      "loss": 2.7276,
+      "step": 17763
+    },
+    {
+      "epoch": 0.9058477542561608,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0009914090340156516,
+      "loss": 2.7366,
+      "step": 17764
+    },
+    {
+      "epoch": 0.9058987477122662,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0009907073733262841,
+      "loss": 2.7777,
+      "step": 17765
+    },
+    {
+      "epoch": 0.9059497411683717,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0009900057800540996,
+      "loss": 2.7457,
+      "step": 17766
+    },
+    {
+      "epoch": 0.906000734624477,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0009893042546318764,
+      "loss": 2.7288,
+      "step": 17767
+    },
+    {
+      "epoch": 0.9060517280805824,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0009886027974923513,
+      "loss": 2.7346,
+      "step": 17768
+    },
+    {
+      "epoch": 0.9061027215366878,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0009879014090682175,
+      "loss": 2.7735,
+      "step": 17769
+    },
+    {
+      "epoch": 0.9061537149927932,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0009872000897921261,
+      "loss": 2.7539,
+      "step": 17770
+    },
+    {
+      "epoch": 0.9062047084488986,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0009864988400966872,
+      "loss": 2.7514,
+      "step": 17771
+    },
+    {
+      "epoch": 0.9062557019050039,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0009857976604144667,
+      "loss": 2.7363,
+      "step": 17772
+    },
+    {
+      "epoch": 0.9063066953611094,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0009850965511779873,
+      "loss": 2.7367,
+      "step": 17773
+    },
+    {
+      "epoch": 0.9063576888172148,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0009843955128197273,
+      "loss": 2.7343,
+      "step": 17774
+    },
+    {
+      "epoch": 0.9064086822733202,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0009836945457721243,
+      "loss": 2.726,
+      "step": 17775
+    },
+    {
+      "epoch": 0.9064596757294255,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0009829936504675686,
+      "loss": 2.7529,
+      "step": 17776
+    },
+    {
+      "epoch": 0.9065106691855309,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0009822928273384081,
+      "loss": 2.7431,
+      "step": 17777
+    },
+    {
+      "epoch": 0.9065616626416363,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0009815920768169457,
+      "loss": 2.7566,
+      "step": 17778
+    },
+    {
+      "epoch": 0.9066126560977417,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0009808913993354397,
+      "loss": 2.7363,
+      "step": 17779
+    },
+    {
+      "epoch": 0.9066636495538472,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.000980190795326103,
+      "loss": 2.7536,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9067146430099525,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0009794902652211035,
+      "loss": 2.7546,
+      "step": 17781
+    },
+    {
+      "epoch": 0.9067656364660579,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0009787898094525626,
+      "loss": 2.7217,
+      "step": 17782
+    },
+    {
+      "epoch": 0.9068166299221633,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0009780894284525578,
+      "loss": 2.7396,
+      "step": 17783
+    },
+    {
+      "epoch": 0.9068676233782687,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0009773891226531187,
+      "loss": 2.7618,
+      "step": 17784
+    },
+    {
+      "epoch": 0.906918616834374,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0009766888924862289,
+      "loss": 2.7352,
+      "step": 17785
+    },
+    {
+      "epoch": 0.9069696102904794,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.000975988738383826,
+      "loss": 2.7675,
+      "step": 17786
+    },
+    {
+      "epoch": 0.9070206037465849,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0009752886607777997,
+      "loss": 2.7182,
+      "step": 17787
+    },
+    {
+      "epoch": 0.9070715972026903,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0009745886600999936,
+      "loss": 2.7422,
+      "step": 17788
+    },
+    {
+      "epoch": 0.9071225906587956,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0009738887367822032,
+      "loss": 2.7475,
+      "step": 17789
+    },
+    {
+      "epoch": 0.907173584114901,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0009731888912561757,
+      "loss": 2.746,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9072245775710064,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0009724891239536118,
+      "loss": 2.7594,
+      "step": 17791
+    },
+    {
+      "epoch": 0.9072755710271118,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0009717894353061627,
+      "loss": 2.7394,
+      "step": 17792
+    },
+    {
+      "epoch": 0.9073265644832171,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.000971089825745432,
+      "loss": 2.7476,
+      "step": 17793
+    },
+    {
+      "epoch": 0.9073775579393226,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0009703902957029731,
+      "loss": 2.7649,
+      "step": 17794
+    },
+    {
+      "epoch": 0.907428551395428,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000969690845610292,
+      "loss": 2.7484,
+      "step": 17795
+    },
+    {
+      "epoch": 0.9074795448515334,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0009689914758988447,
+      "loss": 2.712,
+      "step": 17796
+    },
+    {
+      "epoch": 0.9075305383076387,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0009682921870000374,
+      "loss": 2.7468,
+      "step": 17797
+    },
+    {
+      "epoch": 0.9075815317637441,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0009675929793452263,
+      "loss": 2.729,
+      "step": 17798
+    },
+    {
+      "epoch": 0.9076325252198495,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0009668938533657183,
+      "loss": 2.7611,
+      "step": 17799
+    },
+    {
+      "epoch": 0.9076835186759549,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0009661948094927692,
+      "loss": 2.7098,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9077345121320604,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0009654958481575849,
+      "loss": 2.7495,
+      "step": 17801
+    },
+    {
+      "epoch": 0.9077855055881657,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0009647969697913186,
+      "loss": 2.7237,
+      "step": 17802
+    },
+    {
+      "epoch": 0.9078364990442711,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0009640981748250748,
+      "loss": 2.7295,
+      "step": 17803
+    },
+    {
+      "epoch": 0.9078874925003765,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0009633994636899048,
+      "loss": 2.7668,
+      "step": 17804
+    },
+    {
+      "epoch": 0.9079384859564819,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0009627008368168094,
+      "loss": 2.7125,
+      "step": 17805
+    },
+    {
+      "epoch": 0.9079894794125872,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0009620022946367353,
+      "loss": 2.7207,
+      "step": 17806
+    },
+    {
+      "epoch": 0.9080404728686926,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00096130383758058,
+      "loss": 2.712,
+      "step": 17807
+    },
+    {
+      "epoch": 0.9080914663247981,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0009606054660791861,
+      "loss": 2.7548,
+      "step": 17808
+    },
+    {
+      "epoch": 0.9081424597809035,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0009599071805633446,
+      "loss": 2.7573,
+      "step": 17809
+    },
+    {
+      "epoch": 0.9081934532370088,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0009592089814637921,
+      "loss": 2.7115,
+      "step": 17810
+    },
+    {
+      "epoch": 0.9082444466931142,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0009585108692112144,
+      "loss": 2.7576,
+      "step": 17811
+    },
+    {
+      "epoch": 0.9082954401492196,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0009578128442362413,
+      "loss": 2.7592,
+      "step": 17812
+    },
+    {
+      "epoch": 0.908346433605325,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0009571149069694499,
+      "loss": 2.761,
+      "step": 17813
+    },
+    {
+      "epoch": 0.9083974270614303,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0009564170578413622,
+      "loss": 2.7695,
+      "step": 17814
+    },
+    {
+      "epoch": 0.9084484205175358,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0009557192972824482,
+      "loss": 2.765,
+      "step": 17815
+    },
+    {
+      "epoch": 0.9084994139736412,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0009550216257231204,
+      "loss": 2.7336,
+      "step": 17816
+    },
+    {
+      "epoch": 0.9085504074297466,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0009543240435937381,
+      "loss": 2.7289,
+      "step": 17817
+    },
+    {
+      "epoch": 0.908601400885852,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0009536265513246047,
+      "loss": 2.7126,
+      "step": 17818
+    },
+    {
+      "epoch": 0.9086523943419573,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0009529291493459691,
+      "loss": 2.7113,
+      "step": 17819
+    },
+    {
+      "epoch": 0.9087033877980627,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0009522318380880235,
+      "loss": 2.723,
+      "step": 17820
+    },
+    {
+      "epoch": 0.9087543812541681,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0009515346179809046,
+      "loss": 2.7611,
+      "step": 17821
+    },
+    {
+      "epoch": 0.9088053747102736,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0009508374894546923,
+      "loss": 2.7311,
+      "step": 17822
+    },
+    {
+      "epoch": 0.9088563681663789,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0009501404529394113,
+      "loss": 2.7365,
+      "step": 17823
+    },
+    {
+      "epoch": 0.9089073616224843,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0009494435088650281,
+      "loss": 2.7494,
+      "step": 17824
+    },
+    {
+      "epoch": 0.9089583550785897,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0009487466576614534,
+      "loss": 2.7617,
+      "step": 17825
+    },
+    {
+      "epoch": 0.9090093485346951,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0009480498997585397,
+      "loss": 2.7326,
+      "step": 17826
+    },
+    {
+      "epoch": 0.9090603419908004,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0009473532355860821,
+      "loss": 2.7436,
+      "step": 17827
+    },
+    {
+      "epoch": 0.9091113354469058,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0009466566655738184,
+      "loss": 2.7174,
+      "step": 17828
+    },
+    {
+      "epoch": 0.9091623289030113,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0009459601901514278,
+      "loss": 2.7433,
+      "step": 17829
+    },
+    {
+      "epoch": 0.9092133223591167,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0009452638097485314,
+      "loss": 2.7133,
+      "step": 17830
+    },
+    {
+      "epoch": 0.909264315815222,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0009445675247946916,
+      "loss": 2.7349,
+      "step": 17831
+    },
+    {
+      "epoch": 0.9093153092713274,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0009438713357194122,
+      "loss": 2.7434,
+      "step": 17832
+    },
+    {
+      "epoch": 0.9093663027274328,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0009431752429521372,
+      "loss": 2.7545,
+      "step": 17833
+    },
+    {
+      "epoch": 0.9094172961835382,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0009424792469222516,
+      "loss": 2.725,
+      "step": 17834
+    },
+    {
+      "epoch": 0.9094682896396435,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.000941783348059081,
+      "loss": 2.7283,
+      "step": 17835
+    },
+    {
+      "epoch": 0.909519283095749,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0009410875467918907,
+      "loss": 2.7408,
+      "step": 17836
+    },
+    {
+      "epoch": 0.9095702765518544,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0009403918435498859,
+      "loss": 2.7276,
+      "step": 17837
+    },
+    {
+      "epoch": 0.9096212700079598,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0009396962387622108,
+      "loss": 2.7528,
+      "step": 17838
+    },
+    {
+      "epoch": 0.9096722634640652,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0009390007328579502,
+      "loss": 2.7258,
+      "step": 17839
+    },
+    {
+      "epoch": 0.9097232569201705,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0009383053262661267,
+      "loss": 2.7283,
+      "step": 17840
+    },
+    {
+      "epoch": 0.9097742503762759,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0009376100194157022,
+      "loss": 2.7042,
+      "step": 17841
+    },
+    {
+      "epoch": 0.9098252438323813,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0009369148127355762,
+      "loss": 2.7106,
+      "step": 17842
+    },
+    {
+      "epoch": 0.9098762372884868,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0009362197066545881,
+      "loss": 2.7999,
+      "step": 17843
+    },
+    {
+      "epoch": 0.9099272307445921,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.000935524701601514,
+      "loss": 2.7502,
+      "step": 17844
+    },
+    {
+      "epoch": 0.9099782242006975,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0009348297980050675,
+      "loss": 2.7217,
+      "step": 17845
+    },
+    {
+      "epoch": 0.9100292176568029,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0009341349962938999,
+      "loss": 2.7516,
+      "step": 17846
+    },
+    {
+      "epoch": 0.9100802111129083,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0009334402968966007,
+      "loss": 2.715,
+      "step": 17847
+    },
+    {
+      "epoch": 0.9101312045690136,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0009327457002416948,
+      "loss": 2.7691,
+      "step": 17848
+    },
+    {
+      "epoch": 0.910182198025119,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0009320512067576446,
+      "loss": 2.7261,
+      "step": 17849
+    },
+    {
+      "epoch": 0.9102331914812245,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0009313568168728477,
+      "loss": 2.7676,
+      "step": 17850
+    },
+    {
+      "epoch": 0.9102841849373299,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0009306625310156401,
+      "loss": 2.7372,
+      "step": 17851
+    },
+    {
+      "epoch": 0.9103351783934353,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0009299683496142912,
+      "loss": 2.7394,
+      "step": 17852
+    },
+    {
+      "epoch": 0.9103861718495406,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0009292742730970071,
+      "loss": 2.7432,
+      "step": 17853
+    },
+    {
+      "epoch": 0.910437165305646,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0009285803018919292,
+      "loss": 2.7546,
+      "step": 17854
+    },
+    {
+      "epoch": 0.9104881587617514,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0009278864364271337,
+      "loss": 2.7597,
+      "step": 17855
+    },
+    {
+      "epoch": 0.9105391522178568,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.000927192677130632,
+      "loss": 2.7734,
+      "step": 17856
+    },
+    {
+      "epoch": 0.9105901456739622,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0009264990244303695,
+      "loss": 2.7569,
+      "step": 17857
+    },
+    {
+      "epoch": 0.9106411391300676,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0009258054787542259,
+      "loss": 2.731,
+      "step": 17858
+    },
+    {
+      "epoch": 0.910692132586173,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0009251120405300151,
+      "loss": 2.7548,
+      "step": 17859
+    },
+    {
+      "epoch": 0.9107431260422784,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0009244187101854846,
+      "loss": 2.7172,
+      "step": 17860
+    },
+    {
+      "epoch": 0.9107941194983837,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0009237254881483157,
+      "loss": 2.7144,
+      "step": 17861
+    },
+    {
+      "epoch": 0.9108451129544891,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.000923032374846122,
+      "loss": 2.7453,
+      "step": 17862
+    },
+    {
+      "epoch": 0.9108961064105945,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0009223393707064511,
+      "loss": 2.74,
+      "step": 17863
+    },
+    {
+      "epoch": 0.9109470998667,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0009216464761567827,
+      "loss": 2.7668,
+      "step": 17864
+    },
+    {
+      "epoch": 0.9109980933228053,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0009209536916245287,
+      "loss": 2.7492,
+      "step": 17865
+    },
+    {
+      "epoch": 0.9110490867789107,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0009202610175370338,
+      "loss": 2.7278,
+      "step": 17866
+    },
+    {
+      "epoch": 0.9111000802350161,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0009195684543215739,
+      "loss": 2.7373,
+      "step": 17867
+    },
+    {
+      "epoch": 0.9111510736911215,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0009188760024053566,
+      "loss": 2.7052,
+      "step": 17868
+    },
+    {
+      "epoch": 0.9112020671472268,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0009181836622155217,
+      "loss": 2.7391,
+      "step": 17869
+    },
+    {
+      "epoch": 0.9112530606033322,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0009174914341791391,
+      "loss": 2.7682,
+      "step": 17870
+    },
+    {
+      "epoch": 0.9113040540594377,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0009167993187232091,
+      "loss": 2.7327,
+      "step": 17871
+    },
+    {
+      "epoch": 0.9113550475155431,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0009161073162746644,
+      "loss": 2.7659,
+      "step": 17872
+    },
+    {
+      "epoch": 0.9114060409716485,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0009154154272603663,
+      "loss": 2.709,
+      "step": 17873
+    },
+    {
+      "epoch": 0.9114570344277538,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0009147236521071071,
+      "loss": 2.7566,
+      "step": 17874
+    },
+    {
+      "epoch": 0.9115080278838592,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0009140319912416075,
+      "loss": 2.7546,
+      "step": 17875
+    },
+    {
+      "epoch": 0.9115590213399646,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0009133404450905198,
+      "loss": 2.7657,
+      "step": 17876
+    },
+    {
+      "epoch": 0.91161001479607,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0009126490140804242,
+      "loss": 2.7173,
+      "step": 17877
+    },
+    {
+      "epoch": 0.9116610082521754,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0009119576986378297,
+      "loss": 2.7231,
+      "step": 17878
+    },
+    {
+      "epoch": 0.9117120017082808,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0009112664991891741,
+      "loss": 2.7626,
+      "step": 17879
+    },
+    {
+      "epoch": 0.9117629951643862,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000910575416160825,
+      "loss": 2.7501,
+      "step": 17880
+    },
+    {
+      "epoch": 0.9118139886204916,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0009098844499790764,
+      "loss": 2.7233,
+      "step": 17881
+    },
+    {
+      "epoch": 0.9118649820765969,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0009091936010701511,
+      "loss": 2.7858,
+      "step": 17882
+    },
+    {
+      "epoch": 0.9119159755327023,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0009085028698601989,
+      "loss": 2.7399,
+      "step": 17883
+    },
+    {
+      "epoch": 0.9119669689888077,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0009078122567752989,
+      "loss": 2.7421,
+      "step": 17884
+    },
+    {
+      "epoch": 0.9120179624449132,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0009071217622414545,
+      "loss": 2.7265,
+      "step": 17885
+    },
+    {
+      "epoch": 0.9120689559010186,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0009064313866845984,
+      "loss": 2.7384,
+      "step": 17886
+    },
+    {
+      "epoch": 0.9121199493571239,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.000905741130530588,
+      "loss": 2.7415,
+      "step": 17887
+    },
+    {
+      "epoch": 0.9121709428132293,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0009050509942052087,
+      "loss": 2.7631,
+      "step": 17888
+    },
+    {
+      "epoch": 0.9122219362693347,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0009043609781341713,
+      "loss": 2.7584,
+      "step": 17889
+    },
+    {
+      "epoch": 0.91227292972544,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000903671082743112,
+      "loss": 2.7097,
+      "step": 17890
+    },
+    {
+      "epoch": 0.9123239231815454,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0009029813084575929,
+      "loss": 2.7462,
+      "step": 17891
+    },
+    {
+      "epoch": 0.9123749166376509,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0009022916557031018,
+      "loss": 2.747,
+      "step": 17892
+    },
+    {
+      "epoch": 0.9124259100937563,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.000901602124905051,
+      "loss": 2.7276,
+      "step": 17893
+    },
+    {
+      "epoch": 0.9124769035498617,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0009009127164887778,
+      "loss": 2.7324,
+      "step": 17894
+    },
+    {
+      "epoch": 0.912527897005967,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0009002234308795437,
+      "loss": 2.7433,
+      "step": 17895
+    },
+    {
+      "epoch": 0.9125788904620724,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.000899534268502535,
+      "loss": 2.7437,
+      "step": 17896
+    },
+    {
+      "epoch": 0.9126298839181778,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008988452297828616,
+      "loss": 2.7663,
+      "step": 17897
+    },
+    {
+      "epoch": 0.9126808773742832,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008981563151455572,
+      "loss": 2.7866,
+      "step": 17898
+    },
+    {
+      "epoch": 0.9127318708303886,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008974675250155786,
+      "loss": 2.737,
+      "step": 17899
+    },
+    {
+      "epoch": 0.912782864286494,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0008967788598178068,
+      "loss": 2.7554,
+      "step": 17900
+    },
+    {
+      "epoch": 0.9128338577425994,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0008960903199770447,
+      "loss": 2.7392,
+      "step": 17901
+    },
+    {
+      "epoch": 0.9128848511987048,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0008954019059180185,
+      "loss": 2.7535,
+      "step": 17902
+    },
+    {
+      "epoch": 0.9129358446548101,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008947136180653759,
+      "loss": 2.7646,
+      "step": 17903
+    },
+    {
+      "epoch": 0.9129868381109155,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0008940254568436885,
+      "loss": 2.71,
+      "step": 17904
+    },
+    {
+      "epoch": 0.9130378315670209,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000893337422677448,
+      "loss": 2.7325,
+      "step": 17905
+    },
+    {
+      "epoch": 0.9130888250231264,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0008926495159910688,
+      "loss": 2.7436,
+      "step": 17906
+    },
+    {
+      "epoch": 0.9131398184792318,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0008919617372088856,
+      "loss": 2.7264,
+      "step": 17907
+    },
+    {
+      "epoch": 0.9131908119353371,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0008912740867551559,
+      "loss": 2.7355,
+      "step": 17908
+    },
+    {
+      "epoch": 0.9132418053914425,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0008905865650540563,
+      "loss": 2.7264,
+      "step": 17909
+    },
+    {
+      "epoch": 0.9132927988475479,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0008898991725296854,
+      "loss": 2.7388,
+      "step": 17910
+    },
+    {
+      "epoch": 0.9133437923036533,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0008892119096060605,
+      "loss": 2.7458,
+      "step": 17911
+    },
+    {
+      "epoch": 0.9133947857597586,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008885247767071207,
+      "loss": 2.7405,
+      "step": 17912
+    },
+    {
+      "epoch": 0.9134457792158641,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0008878377742567243,
+      "loss": 2.731,
+      "step": 17913
+    },
+    {
+      "epoch": 0.9134967726719695,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008871509026786479,
+      "loss": 2.749,
+      "step": 17914
+    },
+    {
+      "epoch": 0.9135477661280749,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.000886464162396589,
+      "loss": 2.7295,
+      "step": 17915
+    },
+    {
+      "epoch": 0.9135987595841802,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0008857775538341639,
+      "loss": 2.7346,
+      "step": 17916
+    },
+    {
+      "epoch": 0.9136497530402856,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008850910774149068,
+      "loss": 2.7297,
+      "step": 17917
+    },
+    {
+      "epoch": 0.913700746496391,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0008844047335622707,
+      "loss": 2.7329,
+      "step": 17918
+    },
+    {
+      "epoch": 0.9137517399524964,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0008837185226996267,
+      "loss": 2.7641,
+      "step": 17919
+    },
+    {
+      "epoch": 0.9138027334086019,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008830324452502654,
+      "loss": 2.7562,
+      "step": 17920
+    },
+    {
+      "epoch": 0.9138537268647072,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0008823465016373929,
+      "loss": 2.7347,
+      "step": 17921
+    },
+    {
+      "epoch": 0.9139047203208126,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008816606922841335,
+      "loss": 2.722,
+      "step": 17922
+    },
+    {
+      "epoch": 0.913955713776918,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008809750176135291,
+      "loss": 2.7334,
+      "step": 17923
+    },
+    {
+      "epoch": 0.9140067072330234,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008802894780485388,
+      "loss": 2.7395,
+      "step": 17924
+    },
+    {
+      "epoch": 0.9140577006891287,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0008796040740120373,
+      "loss": 2.7652,
+      "step": 17925
+    },
+    {
+      "epoch": 0.9141086941452341,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0008789188059268165,
+      "loss": 2.7224,
+      "step": 17926
+    },
+    {
+      "epoch": 0.9141596876013396,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008782336742155838,
+      "loss": 2.7522,
+      "step": 17927
+    },
+    {
+      "epoch": 0.914210681057445,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0008775486793009633,
+      "loss": 2.7397,
+      "step": 17928
+    },
+    {
+      "epoch": 0.9142616745135503,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000876863821605494,
+      "loss": 2.718,
+      "step": 17929
+    },
+    {
+      "epoch": 0.9143126679696557,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000876179101551631,
+      "loss": 2.753,
+      "step": 17930
+    },
+    {
+      "epoch": 0.9143636614257611,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0008754945195617432,
+      "loss": 2.7478,
+      "step": 17931
+    },
+    {
+      "epoch": 0.9144146548818665,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008748100760581159,
+      "loss": 2.7377,
+      "step": 17932
+    },
+    {
+      "epoch": 0.9144656483379718,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0008741257714629478,
+      "loss": 2.766,
+      "step": 17933
+    },
+    {
+      "epoch": 0.9145166417940773,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008734416061983528,
+      "loss": 2.734,
+      "step": 17934
+    },
+    {
+      "epoch": 0.9145676352501827,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0008727575806863579,
+      "loss": 2.7491,
+      "step": 17935
+    },
+    {
+      "epoch": 0.9146186287062881,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008720736953489048,
+      "loss": 2.7461,
+      "step": 17936
+    },
+    {
+      "epoch": 0.9146696221623934,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0008713899506078482,
+      "loss": 2.7455,
+      "step": 17937
+    },
+    {
+      "epoch": 0.9147206156184988,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008707063468849563,
+      "loss": 2.735,
+      "step": 17938
+    },
+    {
+      "epoch": 0.9147716090746042,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0008700228846019098,
+      "loss": 2.7658,
+      "step": 17939
+    },
+    {
+      "epoch": 0.9148226025307096,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008693395641803031,
+      "loss": 2.7152,
+      "step": 17940
+    },
+    {
+      "epoch": 0.9148735959868151,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0008686563860416423,
+      "loss": 2.7354,
+      "step": 17941
+    },
+    {
+      "epoch": 0.9149245894429204,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0008679733506073462,
+      "loss": 2.763,
+      "step": 17942
+    },
+    {
+      "epoch": 0.9149755828990258,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.000867290458298745,
+      "loss": 2.7384,
+      "step": 17943
+    },
+    {
+      "epoch": 0.9150265763551312,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0008666077095370815,
+      "loss": 2.7235,
+      "step": 17944
+    },
+    {
+      "epoch": 0.9150775698112366,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.000865925104743509,
+      "loss": 2.7307,
+      "step": 17945
+    },
+    {
+      "epoch": 0.9151285632673419,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000865242644339093,
+      "loss": 2.7322,
+      "step": 17946
+    },
+    {
+      "epoch": 0.9151795567234473,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0008645603287448083,
+      "loss": 2.7067,
+      "step": 17947
+    },
+    {
+      "epoch": 0.9152305501795528,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008638781583815427,
+      "loss": 2.7617,
+      "step": 17948
+    },
+    {
+      "epoch": 0.9152815436356582,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0008631961336700925,
+      "loss": 2.7495,
+      "step": 17949
+    },
+    {
+      "epoch": 0.9153325370917635,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0008625142550311647,
+      "loss": 2.7298,
+      "step": 17950
+    },
+    {
+      "epoch": 0.9153835305478689,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000861832522885376,
+      "loss": 2.7475,
+      "step": 17951
+    },
+    {
+      "epoch": 0.9154345240039743,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008611509376532541,
+      "loss": 2.7401,
+      "step": 17952
+    },
+    {
+      "epoch": 0.9154855174600797,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0008604694997552345,
+      "loss": 2.7372,
+      "step": 17953
+    },
+    {
+      "epoch": 0.915536510916185,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0008597882096116622,
+      "loss": 2.7594,
+      "step": 17954
+    },
+    {
+      "epoch": 0.9155875043722905,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0008591070676427908,
+      "loss": 2.7475,
+      "step": 17955
+    },
+    {
+      "epoch": 0.9156384978283959,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.000858426074268784,
+      "loss": 2.7405,
+      "step": 17956
+    },
+    {
+      "epoch": 0.9156894912845013,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0008577452299097117,
+      "loss": 2.7268,
+      "step": 17957
+    },
+    {
+      "epoch": 0.9157404847406067,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008570645349855535,
+      "loss": 2.7578,
+      "step": 17958
+    },
+    {
+      "epoch": 0.915791478196712,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0008563839899161958,
+      "loss": 2.7554,
+      "step": 17959
+    },
+    {
+      "epoch": 0.9158424716528174,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0008557035951214333,
+      "loss": 2.7484,
+      "step": 17960
+    },
+    {
+      "epoch": 0.9158934651089228,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0008550233510209675,
+      "loss": 2.7346,
+      "step": 17961
+    },
+    {
+      "epoch": 0.9159444585650283,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008543432580344075,
+      "loss": 2.7179,
+      "step": 17962
+    },
+    {
+      "epoch": 0.9159954520211336,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0008536633165812682,
+      "loss": 2.7557,
+      "step": 17963
+    },
+    {
+      "epoch": 0.916046445477239,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0008529835270809723,
+      "loss": 2.7576,
+      "step": 17964
+    },
+    {
+      "epoch": 0.9160974389333444,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008523038899528476,
+      "loss": 2.7261,
+      "step": 17965
+    },
+    {
+      "epoch": 0.9161484323894498,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000851624405616129,
+      "loss": 2.7257,
+      "step": 17966
+    },
+    {
+      "epoch": 0.9161994258455551,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0008509450744899558,
+      "loss": 2.7503,
+      "step": 17967
+    },
+    {
+      "epoch": 0.9162504193016605,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008502658969933745,
+      "loss": 2.7329,
+      "step": 17968
+    },
+    {
+      "epoch": 0.916301412757766,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008495868735453352,
+      "loss": 2.7428,
+      "step": 17969
+    },
+    {
+      "epoch": 0.9163524062138714,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008489080045646938,
+      "loss": 2.759,
+      "step": 17970
+    },
+    {
+      "epoch": 0.9164033996699767,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008482292904702109,
+      "loss": 2.739,
+      "step": 17971
+    },
+    {
+      "epoch": 0.9164543931260821,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0008475507316805515,
+      "loss": 2.7386,
+      "step": 17972
+    },
+    {
+      "epoch": 0.9165053865821875,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0008468723286142848,
+      "loss": 2.7233,
+      "step": 17973
+    },
+    {
+      "epoch": 0.9165563800382929,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0008461940816898837,
+      "loss": 2.7179,
+      "step": 17974
+    },
+    {
+      "epoch": 0.9166073734943982,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008455159913257249,
+      "loss": 2.6958,
+      "step": 17975
+    },
+    {
+      "epoch": 0.9166583669505037,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008448380579400889,
+      "loss": 2.7248,
+      "step": 17976
+    },
+    {
+      "epoch": 0.9167093604066091,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008441602819511591,
+      "loss": 2.7334,
+      "step": 17977
+    },
+    {
+      "epoch": 0.9167603538627145,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008434826637770216,
+      "loss": 2.7387,
+      "step": 17978
+    },
+    {
+      "epoch": 0.9168113473188199,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008428052038356651,
+      "loss": 2.7396,
+      "step": 17979
+    },
+    {
+      "epoch": 0.9168623407749252,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008421279025449815,
+      "loss": 2.7497,
+      "step": 17980
+    },
+    {
+      "epoch": 0.9169133342310306,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008414507603227642,
+      "loss": 2.7365,
+      "step": 17981
+    },
+    {
+      "epoch": 0.916964327687136,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008407737775867082,
+      "loss": 2.709,
+      "step": 17982
+    },
+    {
+      "epoch": 0.9170153211432415,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0008400969547544097,
+      "loss": 2.7275,
+      "step": 17983
+    },
+    {
+      "epoch": 0.9170663145993468,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0008394202922433687,
+      "loss": 2.7368,
+      "step": 17984
+    },
+    {
+      "epoch": 0.9171173080554522,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.000838743790470984,
+      "loss": 2.7319,
+      "step": 17985
+    },
+    {
+      "epoch": 0.9171683015115576,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008380674498545555,
+      "loss": 2.7119,
+      "step": 17986
+    },
+    {
+      "epoch": 0.917219294967663,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008373912708112845,
+      "loss": 2.7594,
+      "step": 17987
+    },
+    {
+      "epoch": 0.9172702884237683,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008367152537582714,
+      "loss": 2.7486,
+      "step": 17988
+    },
+    {
+      "epoch": 0.9173212818798737,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008360393991125185,
+      "loss": 2.7144,
+      "step": 17989
+    },
+    {
+      "epoch": 0.9173722753359792,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0008353637072909266,
+      "loss": 2.725,
+      "step": 17990
+    },
+    {
+      "epoch": 0.9174232687920846,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008346881787102962,
+      "loss": 2.7474,
+      "step": 17991
+    },
+    {
+      "epoch": 0.91747426224819,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008340128137873271,
+      "loss": 2.7349,
+      "step": 17992
+    },
+    {
+      "epoch": 0.9175252557042953,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0008333376129386188,
+      "loss": 2.774,
+      "step": 17993
+    },
+    {
+      "epoch": 0.9175762491604007,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008326625765806688,
+      "loss": 2.7512,
+      "step": 17994
+    },
+    {
+      "epoch": 0.9176272426165061,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008319877051298734,
+      "loss": 2.7459,
+      "step": 17995
+    },
+    {
+      "epoch": 0.9176782360726115,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0008313129990025273,
+      "loss": 2.7299,
+      "step": 17996
+    },
+    {
+      "epoch": 0.9177292295287169,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008306384586148233,
+      "loss": 2.7364,
+      "step": 17997
+    },
+    {
+      "epoch": 0.9177802229848223,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0008299640843828515,
+      "loss": 2.7408,
+      "step": 17998
+    },
+    {
+      "epoch": 0.9178312164409277,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0008292898767226002,
+      "loss": 2.6971,
+      "step": 17999
+    },
+    {
+      "epoch": 0.9178822098970331,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0008286158360499541,
+      "loss": 2.7165,
+      "step": 18000
+    },
+    {
+      "epoch": 0.9179332033531384,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008279419627806955,
+      "loss": 2.7176,
+      "step": 18001
+    },
+    {
+      "epoch": 0.9179841968092438,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008272682573305037,
+      "loss": 2.7461,
+      "step": 18002
+    },
+    {
+      "epoch": 0.9180351902653492,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0008265947201149535,
+      "loss": 2.7556,
+      "step": 18003
+    },
+    {
+      "epoch": 0.9180861837214547,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0008259213515495166,
+      "loss": 2.7507,
+      "step": 18004
+    },
+    {
+      "epoch": 0.91813717717756,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000825248152049561,
+      "loss": 2.7353,
+      "step": 18005
+    },
+    {
+      "epoch": 0.9181881706336654,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0008245751220303495,
+      "loss": 2.711,
+      "step": 18006
+    },
+    {
+      "epoch": 0.9182391640897708,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0008239022619070411,
+      "loss": 2.7429,
+      "step": 18007
+    },
+    {
+      "epoch": 0.9182901575458762,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008232295720946894,
+      "loss": 2.7228,
+      "step": 18008
+    },
+    {
+      "epoch": 0.9183411510019815,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0008225570530082438,
+      "loss": 2.7423,
+      "step": 18009
+    },
+    {
+      "epoch": 0.9183921444580869,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008218847050625476,
+      "loss": 2.7241,
+      "step": 18010
+    },
+    {
+      "epoch": 0.9184431379141924,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0008212125286723384,
+      "loss": 2.7341,
+      "step": 18011
+    },
+    {
+      "epoch": 0.9184941313702978,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008205405242522487,
+      "loss": 2.7376,
+      "step": 18012
+    },
+    {
+      "epoch": 0.9185451248264032,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000819868692216805,
+      "loss": 2.7447,
+      "step": 18013
+    },
+    {
+      "epoch": 0.9185961182825085,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008191970329804268,
+      "loss": 2.7422,
+      "step": 18014
+    },
+    {
+      "epoch": 0.9186471117386139,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008185255469574271,
+      "loss": 2.7331,
+      "step": 18015
+    },
+    {
+      "epoch": 0.9186981051947193,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0008178542345620117,
+      "loss": 2.7194,
+      "step": 18016
+    },
+    {
+      "epoch": 0.9187490986508247,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0008171830962082807,
+      "loss": 2.7168,
+      "step": 18017
+    },
+    {
+      "epoch": 0.9188000921069301,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0008165121323102261,
+      "loss": 2.737,
+      "step": 18018
+    },
+    {
+      "epoch": 0.9188510855630355,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008158413432817311,
+      "loss": 2.7156,
+      "step": 18019
+    },
+    {
+      "epoch": 0.9189020790191409,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008151707295365725,
+      "loss": 2.6998,
+      "step": 18020
+    },
+    {
+      "epoch": 0.9189530724752463,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000814500291488419,
+      "loss": 2.7332,
+      "step": 18021
+    },
+    {
+      "epoch": 0.9190040659313516,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0008138300295508303,
+      "loss": 2.718,
+      "step": 18022
+    },
+    {
+      "epoch": 0.919055059387457,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0008131599441372575,
+      "loss": 2.7417,
+      "step": 18023
+    },
+    {
+      "epoch": 0.9191060528435624,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0008124900356610425,
+      "loss": 2.7051,
+      "step": 18024
+    },
+    {
+      "epoch": 0.9191570462996679,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0008118203045354195,
+      "loss": 2.7249,
+      "step": 18025
+    },
+    {
+      "epoch": 0.9192080397557733,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0008111507511735117,
+      "loss": 2.7314,
+      "step": 18026
+    },
+    {
+      "epoch": 0.9192590332118786,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008104813759883334,
+      "loss": 2.729,
+      "step": 18027
+    },
+    {
+      "epoch": 0.919310026667984,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008098121793927882,
+      "loss": 2.7386,
+      "step": 18028
+    },
+    {
+      "epoch": 0.9193610201240894,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008091431617996714,
+      "loss": 2.7525,
+      "step": 18029
+    },
+    {
+      "epoch": 0.9194120135801948,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0008084743236216658,
+      "loss": 2.7372,
+      "step": 18030
+    },
+    {
+      "epoch": 0.9194630070363001,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008078056652713446,
+      "loss": 2.7013,
+      "step": 18031
+    },
+    {
+      "epoch": 0.9195140004924056,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0008071371871611697,
+      "loss": 2.7303,
+      "step": 18032
+    },
+    {
+      "epoch": 0.919564993948511,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008064688897034924,
+      "loss": 2.7522,
+      "step": 18033
+    },
+    {
+      "epoch": 0.9196159874046164,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.000805800773310552,
+      "loss": 2.7353,
+      "step": 18034
+    },
+    {
+      "epoch": 0.9196669808607217,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008051328383944761,
+      "loss": 2.7072,
+      "step": 18035
+    },
+    {
+      "epoch": 0.9197179743168271,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008044650853672804,
+      "loss": 2.7228,
+      "step": 18036
+    },
+    {
+      "epoch": 0.9197689677729325,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.000803797514640869,
+      "loss": 2.7151,
+      "step": 18037
+    },
+    {
+      "epoch": 0.9198199612290379,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0008031301266270325,
+      "loss": 2.7214,
+      "step": 18038
+    },
+    {
+      "epoch": 0.9198709546851433,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0008024629217374497,
+      "loss": 2.7347,
+      "step": 18039
+    },
+    {
+      "epoch": 0.9199219481412487,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008017959003836861,
+      "loss": 2.7496,
+      "step": 18040
+    },
+    {
+      "epoch": 0.9199729415973541,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0008011290629771939,
+      "loss": 2.7367,
+      "step": 18041
+    },
+    {
+      "epoch": 0.9200239350534595,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008004624099293121,
+      "loss": 2.7286,
+      "step": 18042
+    },
+    {
+      "epoch": 0.9200749285095648,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007997959416512655,
+      "loss": 2.7338,
+      "step": 18043
+    },
+    {
+      "epoch": 0.9201259219656702,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007991296585541652,
+      "loss": 2.7398,
+      "step": 18044
+    },
+    {
+      "epoch": 0.9201769154217756,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007984635610490085,
+      "loss": 2.7421,
+      "step": 18045
+    },
+    {
+      "epoch": 0.9202279088778811,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0007977976495466775,
+      "loss": 2.7347,
+      "step": 18046
+    },
+    {
+      "epoch": 0.9202789023339865,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007971319244579399,
+      "loss": 2.7581,
+      "step": 18047
+    },
+    {
+      "epoch": 0.9203298957900918,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0007964663861934482,
+      "loss": 2.687,
+      "step": 18048
+    },
+    {
+      "epoch": 0.9203808892461972,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007958010351637404,
+      "loss": 2.7476,
+      "step": 18049
+    },
+    {
+      "epoch": 0.9204318827023026,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007951358717792379,
+      "loss": 2.7275,
+      "step": 18050
+    },
+    {
+      "epoch": 0.920482876158408,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007944708964502474,
+      "loss": 2.7378,
+      "step": 18051
+    },
+    {
+      "epoch": 0.9205338696145133,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000793806109586958,
+      "loss": 2.722,
+      "step": 18052
+    },
+    {
+      "epoch": 0.9205848630706188,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0007931415115994449,
+      "loss": 2.7324,
+      "step": 18053
+    },
+    {
+      "epoch": 0.9206358565267242,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007924771028976652,
+      "loss": 2.6973,
+      "step": 18054
+    },
+    {
+      "epoch": 0.9206868499828296,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007918128838914593,
+      "loss": 2.7407,
+      "step": 18055
+    },
+    {
+      "epoch": 0.9207378434389349,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007911488549905506,
+      "loss": 2.7248,
+      "step": 18056
+    },
+    {
+      "epoch": 0.9207888368950403,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007904850166045464,
+      "loss": 2.7388,
+      "step": 18057
+    },
+    {
+      "epoch": 0.9208398303511457,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007898213691429348,
+      "loss": 2.735,
+      "step": 18058
+    },
+    {
+      "epoch": 0.9208908238072511,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007891579130150872,
+      "loss": 2.711,
+      "step": 18059
+    },
+    {
+      "epoch": 0.9209418172633566,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0007884946486302562,
+      "loss": 2.7305,
+      "step": 18060
+    },
+    {
+      "epoch": 0.9209928107194619,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0007878315763975777,
+      "loss": 2.722,
+      "step": 18061
+    },
+    {
+      "epoch": 0.9210438041755673,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007871686967260667,
+      "loss": 2.7196,
+      "step": 18062
+    },
+    {
+      "epoch": 0.9210947976316727,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007865060100246217,
+      "loss": 2.7441,
+      "step": 18063
+    },
+    {
+      "epoch": 0.921145791087778,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0007858435167020202,
+      "loss": 2.7249,
+      "step": 18064
+    },
+    {
+      "epoch": 0.9211967845438834,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0007851812171669222,
+      "loss": 2.7381,
+      "step": 18065
+    },
+    {
+      "epoch": 0.9212477779999888,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007845191118278672,
+      "loss": 2.7265,
+      "step": 18066
+    },
+    {
+      "epoch": 0.9212987714560943,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007838572010932745,
+      "loss": 2.7363,
+      "step": 18067
+    },
+    {
+      "epoch": 0.9213497649121997,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007831954853714443,
+      "loss": 2.738,
+      "step": 18068
+    },
+    {
+      "epoch": 0.921400758368305,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007825339650705561,
+      "loss": 2.7354,
+      "step": 18069
+    },
+    {
+      "epoch": 0.9214517518244104,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007818726405986687,
+      "loss": 2.7471,
+      "step": 18070
+    },
+    {
+      "epoch": 0.9215027452805158,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007812115123637206,
+      "loss": 2.7586,
+      "step": 18071
+    },
+    {
+      "epoch": 0.9215537387366212,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0007805505807735282,
+      "loss": 2.7457,
+      "step": 18072
+    },
+    {
+      "epoch": 0.9216047321927265,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0007798898462357884,
+      "loss": 2.7396,
+      "step": 18073
+    },
+    {
+      "epoch": 0.921655725648832,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007792293091580745,
+      "loss": 2.7329,
+      "step": 18074
+    },
+    {
+      "epoch": 0.9217067191049374,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007785689699478396,
+      "loss": 2.7412,
+      "step": 18075
+    },
+    {
+      "epoch": 0.9217577125610428,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007779088290124132,
+      "loss": 2.7338,
+      "step": 18076
+    },
+    {
+      "epoch": 0.9218087060171481,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007772488867590046,
+      "loss": 2.738,
+      "step": 18077
+    },
+    {
+      "epoch": 0.9218596994732535,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007765891435946986,
+      "loss": 2.7391,
+      "step": 18078
+    },
+    {
+      "epoch": 0.9219106929293589,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000775929599926458,
+      "loss": 2.7239,
+      "step": 18079
+    },
+    {
+      "epoch": 0.9219616863854643,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0007752702561611225,
+      "loss": 2.7306,
+      "step": 18080
+    },
+    {
+      "epoch": 0.9220126798415698,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007746111127054085,
+      "loss": 2.7412,
+      "step": 18081
+    },
+    {
+      "epoch": 0.9220636732976751,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007739521699659089,
+      "loss": 2.7286,
+      "step": 18082
+    },
+    {
+      "epoch": 0.9221146667537805,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007732934283490926,
+      "loss": 2.7376,
+      "step": 18083
+    },
+    {
+      "epoch": 0.9221656602098859,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007726348882613046,
+      "loss": 2.7217,
+      "step": 18084
+    },
+    {
+      "epoch": 0.9222166536659913,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007719765501087652,
+      "loss": 2.7218,
+      "step": 18085
+    },
+    {
+      "epoch": 0.9222676471220966,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007713184142975711,
+      "loss": 2.7271,
+      "step": 18086
+    },
+    {
+      "epoch": 0.922318640578202,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000770660481233693,
+      "loss": 2.7607,
+      "step": 18087
+    },
+    {
+      "epoch": 0.9223696340343075,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007700027513229768,
+      "loss": 2.736,
+      "step": 18088
+    },
+    {
+      "epoch": 0.9224206274904129,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0007693452249711441,
+      "loss": 2.7132,
+      "step": 18089
+    },
+    {
+      "epoch": 0.9224716209465182,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007686879025837898,
+      "loss": 2.7506,
+      "step": 18090
+    },
+    {
+      "epoch": 0.9225226144026236,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0007680307845663838,
+      "loss": 2.7014,
+      "step": 18091
+    },
+    {
+      "epoch": 0.922573607858729,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0007673738713242684,
+      "loss": 2.7311,
+      "step": 18092
+    },
+    {
+      "epoch": 0.9226246013148344,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007667171632626617,
+      "loss": 2.7418,
+      "step": 18093
+    },
+    {
+      "epoch": 0.9226755947709397,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0007660606607866543,
+      "loss": 2.731,
+      "step": 18094
+    },
+    {
+      "epoch": 0.9227265882270452,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0007654043643012092,
+      "loss": 2.7379,
+      "step": 18095
+    },
+    {
+      "epoch": 0.9227775816831506,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007647482742111633,
+      "loss": 2.7234,
+      "step": 18096
+    },
+    {
+      "epoch": 0.922828575139256,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007640923909212266,
+      "loss": 2.7185,
+      "step": 18097
+    },
+    {
+      "epoch": 0.9228795685953614,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007634367148359797,
+      "loss": 2.7502,
+      "step": 18098
+    },
+    {
+      "epoch": 0.9229305620514667,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007627812463598775,
+      "loss": 2.7708,
+      "step": 18099
+    },
+    {
+      "epoch": 0.9229815555075721,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007621259858972451,
+      "loss": 2.7434,
+      "step": 18100
+    },
+    {
+      "epoch": 0.9230325489636775,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007614709338522812,
+      "loss": 2.7276,
+      "step": 18101
+    },
+    {
+      "epoch": 0.923083542419783,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007608160906290537,
+      "loss": 2.7288,
+      "step": 18102
+    },
+    {
+      "epoch": 0.9231345358758883,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0007601614566315034,
+      "loss": 2.7471,
+      "step": 18103
+    },
+    {
+      "epoch": 0.9231855293319937,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007595070322634412,
+      "loss": 2.7683,
+      "step": 18104
+    },
+    {
+      "epoch": 0.9232365227880991,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000758852817928549,
+      "loss": 2.7269,
+      "step": 18105
+    },
+    {
+      "epoch": 0.9232875162442045,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007581988140303791,
+      "loss": 2.7546,
+      "step": 18106
+    },
+    {
+      "epoch": 0.9233385097003098,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000757545020972354,
+      "loss": 2.7424,
+      "step": 18107
+    },
+    {
+      "epoch": 0.9233895031564152,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007568914391577659,
+      "loss": 2.7386,
+      "step": 18108
+    },
+    {
+      "epoch": 0.9234404966125207,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007562380689897772,
+      "loss": 2.7279,
+      "step": 18109
+    },
+    {
+      "epoch": 0.9234914900686261,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007555849108714192,
+      "loss": 2.7061,
+      "step": 18110
+    },
+    {
+      "epoch": 0.9235424835247315,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.000754931965205593,
+      "loss": 2.7322,
+      "step": 18111
+    },
+    {
+      "epoch": 0.9235934769808368,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0007542792323950678,
+      "loss": 2.7616,
+      "step": 18112
+    },
+    {
+      "epoch": 0.9236444704369422,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007536267128424823,
+      "loss": 2.7271,
+      "step": 18113
+    },
+    {
+      "epoch": 0.9236954638930476,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0007529744069503432,
+      "loss": 2.7394,
+      "step": 18114
+    },
+    {
+      "epoch": 0.923746457349153,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007523223151210259,
+      "loss": 2.7444,
+      "step": 18115
+    },
+    {
+      "epoch": 0.9237974508052584,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007516704377567726,
+      "loss": 2.7344,
+      "step": 18116
+    },
+    {
+      "epoch": 0.9238484442613638,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0007510187752596949,
+      "loss": 2.745,
+      "step": 18117
+    },
+    {
+      "epoch": 0.9238994377174692,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007503673280317706,
+      "loss": 2.7207,
+      "step": 18118
+    },
+    {
+      "epoch": 0.9239504311735746,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.000749716096474845,
+      "loss": 2.7197,
+      "step": 18119
+    },
+    {
+      "epoch": 0.9240014246296799,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007490650809906308,
+      "loss": 2.7632,
+      "step": 18120
+    },
+    {
+      "epoch": 0.9240524180857853,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007484142819807062,
+      "loss": 2.7354,
+      "step": 18121
+    },
+    {
+      "epoch": 0.9241034115418907,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007477636998465178,
+      "loss": 2.7506,
+      "step": 18122
+    },
+    {
+      "epoch": 0.9241544049979962,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007471133349893768,
+      "loss": 2.7086,
+      "step": 18123
+    },
+    {
+      "epoch": 0.9242053984541015,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0007464631878104612,
+      "loss": 2.7217,
+      "step": 18124
+    },
+    {
+      "epoch": 0.9242563919102069,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007458132587108137,
+      "loss": 2.7603,
+      "step": 18125
+    },
+    {
+      "epoch": 0.9243073853663123,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007451635480913441,
+      "loss": 2.742,
+      "step": 18126
+    },
+    {
+      "epoch": 0.9243583788224177,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007445140563528263,
+      "loss": 2.7085,
+      "step": 18127
+    },
+    {
+      "epoch": 0.924409372278523,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007438647838958994,
+      "loss": 2.7265,
+      "step": 18128
+    },
+    {
+      "epoch": 0.9244603657346284,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0007432157311210668,
+      "loss": 2.737,
+      "step": 18129
+    },
+    {
+      "epoch": 0.9245113591907339,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007425668984286975,
+      "loss": 2.7208,
+      "step": 18130
+    },
+    {
+      "epoch": 0.9245623526468393,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007419182862190241,
+      "loss": 2.7177,
+      "step": 18131
+    },
+    {
+      "epoch": 0.9246133461029447,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0007412698948921428,
+      "loss": 2.7142,
+      "step": 18132
+    },
+    {
+      "epoch": 0.92466433955905,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007406217248480139,
+      "loss": 2.7204,
+      "step": 18133
+    },
+    {
+      "epoch": 0.9247153330151554,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007399737764864618,
+      "loss": 2.7124,
+      "step": 18134
+    },
+    {
+      "epoch": 0.9247663264712608,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007393260502071733,
+      "loss": 2.7212,
+      "step": 18135
+    },
+    {
+      "epoch": 0.9248173199273662,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007386785464096985,
+      "loss": 2.741,
+      "step": 18136
+    },
+    {
+      "epoch": 0.9248683133834716,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00073803126549345,
+      "loss": 2.7003,
+      "step": 18137
+    },
+    {
+      "epoch": 0.924919306839577,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007373842078577038,
+      "loss": 2.726,
+      "step": 18138
+    },
+    {
+      "epoch": 0.9249703002956824,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000736737373901597,
+      "loss": 2.7387,
+      "step": 18139
+    },
+    {
+      "epoch": 0.9250212937517878,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007360907640241297,
+      "loss": 2.726,
+      "step": 18140
+    },
+    {
+      "epoch": 0.9250722872078931,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007354443786241629,
+      "loss": 2.7331,
+      "step": 18141
+    },
+    {
+      "epoch": 0.9251232806639985,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0007347982181004203,
+      "loss": 2.7305,
+      "step": 18142
+    },
+    {
+      "epoch": 0.9251742741201039,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007341522828514855,
+      "loss": 2.7564,
+      "step": 18143
+    },
+    {
+      "epoch": 0.9252252675762094,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.000733506573275804,
+      "loss": 2.7388,
+      "step": 18144
+    },
+    {
+      "epoch": 0.9252762610323148,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007328610897716819,
+      "loss": 2.7164,
+      "step": 18145
+    },
+    {
+      "epoch": 0.9253272544884201,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0007322158327372861,
+      "loss": 2.7215,
+      "step": 18146
+    },
+    {
+      "epoch": 0.9253782479445255,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007315708025706432,
+      "loss": 2.704,
+      "step": 18147
+    },
+    {
+      "epoch": 0.9254292414006309,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007309259996696405,
+      "loss": 2.756,
+      "step": 18148
+    },
+    {
+      "epoch": 0.9254802348567363,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007302814244320246,
+      "loss": 2.7103,
+      "step": 18149
+    },
+    {
+      "epoch": 0.9255312283128416,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007296370772554022,
+      "loss": 2.7263,
+      "step": 18150
+    },
+    {
+      "epoch": 0.9255822217689471,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007289929585372388,
+      "loss": 2.7253,
+      "step": 18151
+    },
+    {
+      "epoch": 0.9256332152250525,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007283490686748591,
+      "loss": 2.7192,
+      "step": 18152
+    },
+    {
+      "epoch": 0.9256842086811579,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007277054080654468,
+      "loss": 2.7476,
+      "step": 18153
+    },
+    {
+      "epoch": 0.9257352021372632,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007270619771060443,
+      "loss": 2.7419,
+      "step": 18154
+    },
+    {
+      "epoch": 0.9257861955933686,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007264187761935519,
+      "loss": 2.741,
+      "step": 18155
+    },
+    {
+      "epoch": 0.925837189049474,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007257758057247286,
+      "loss": 2.7486,
+      "step": 18156
+    },
+    {
+      "epoch": 0.9258881825055794,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0007251330660961904,
+      "loss": 2.7283,
+      "step": 18157
+    },
+    {
+      "epoch": 0.9259391759616848,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0007244905577044115,
+      "loss": 2.7266,
+      "step": 18158
+    },
+    {
+      "epoch": 0.9259901694177902,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007238482809457234,
+      "loss": 2.7438,
+      "step": 18159
+    },
+    {
+      "epoch": 0.9260411628738956,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000723206236216315,
+      "loss": 2.7279,
+      "step": 18160
+    },
+    {
+      "epoch": 0.926092156330001,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007225644239122311,
+      "loss": 2.7189,
+      "step": 18161
+    },
+    {
+      "epoch": 0.9261431497861063,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007219228444293744,
+      "loss": 2.7199,
+      "step": 18162
+    },
+    {
+      "epoch": 0.9261941432422117,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007212814981635032,
+      "loss": 2.6958,
+      "step": 18163
+    },
+    {
+      "epoch": 0.9262451366983171,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.000720640385510232,
+      "loss": 2.7511,
+      "step": 18164
+    },
+    {
+      "epoch": 0.9262961301544226,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007199995068650311,
+      "loss": 2.7336,
+      "step": 18165
+    },
+    {
+      "epoch": 0.926347123610528,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007193588626232272,
+      "loss": 2.704,
+      "step": 18166
+    },
+    {
+      "epoch": 0.9263981170666333,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0007187184531800015,
+      "loss": 2.7366,
+      "step": 18167
+    },
+    {
+      "epoch": 0.9264491105227387,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007180782789303905,
+      "loss": 2.733,
+      "step": 18168
+    },
+    {
+      "epoch": 0.9265001039788441,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007174383402692859,
+      "loss": 2.7193,
+      "step": 18169
+    },
+    {
+      "epoch": 0.9265510974349495,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007167986375914346,
+      "loss": 2.7304,
+      "step": 18170
+    },
+    {
+      "epoch": 0.9266020908910548,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007161591712914368,
+      "loss": 2.71,
+      "step": 18171
+    },
+    {
+      "epoch": 0.9266530843471603,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007155199417637479,
+      "loss": 2.7191,
+      "step": 18172
+    },
+    {
+      "epoch": 0.9267040778032657,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.000714880949402676,
+      "loss": 2.7569,
+      "step": 18173
+    },
+    {
+      "epoch": 0.9267550712593711,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0007142421946023847,
+      "loss": 2.7383,
+      "step": 18174
+    },
+    {
+      "epoch": 0.9268060647154764,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0007136036777568893,
+      "loss": 2.7536,
+      "step": 18175
+    },
+    {
+      "epoch": 0.9268570581715818,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007129653992600593,
+      "loss": 2.7356,
+      "step": 18176
+    },
+    {
+      "epoch": 0.9269080516276872,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007123273595056166,
+      "loss": 2.7326,
+      "step": 18177
+    },
+    {
+      "epoch": 0.9269590450837926,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007116895588871368,
+      "loss": 2.7344,
+      "step": 18178
+    },
+    {
+      "epoch": 0.927010038539898,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007110519977980474,
+      "loss": 2.7187,
+      "step": 18179
+    },
+    {
+      "epoch": 0.9270610319960034,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0007104146766316279,
+      "loss": 2.7249,
+      "step": 18180
+    },
+    {
+      "epoch": 0.9271120254521088,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007097775957810091,
+      "loss": 2.7319,
+      "step": 18181
+    },
+    {
+      "epoch": 0.9271630189082142,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000709140755639176,
+      "loss": 2.6954,
+      "step": 18182
+    },
+    {
+      "epoch": 0.9272140123643196,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007085041565989627,
+      "loss": 2.737,
+      "step": 18183
+    },
+    {
+      "epoch": 0.9272650058204249,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0007078677990530558,
+      "loss": 2.7083,
+      "step": 18184
+    },
+    {
+      "epoch": 0.9273159992765303,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.000707231683393992,
+      "loss": 2.7212,
+      "step": 18185
+    },
+    {
+      "epoch": 0.9273669927326358,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0007065958100141607,
+      "loss": 2.7104,
+      "step": 18186
+    },
+    {
+      "epoch": 0.9274179861887412,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007059601793057992,
+      "loss": 2.6937,
+      "step": 18187
+    },
+    {
+      "epoch": 0.9274689796448465,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007053247916609971,
+      "loss": 2.7257,
+      "step": 18188
+    },
+    {
+      "epoch": 0.9275199731009519,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007046896474716931,
+      "loss": 2.6963,
+      "step": 18189
+    },
+    {
+      "epoch": 0.9275709665570573,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007040547471296763,
+      "loss": 2.7112,
+      "step": 18190
+    },
+    {
+      "epoch": 0.9276219600131627,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007034200910265853,
+      "loss": 2.7417,
+      "step": 18191
+    },
+    {
+      "epoch": 0.927672953469268,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0007027856795539075,
+      "loss": 2.7193,
+      "step": 18192
+    },
+    {
+      "epoch": 0.9277239469253735,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007021515131029803,
+      "loss": 2.7139,
+      "step": 18193
+    },
+    {
+      "epoch": 0.9277749403814789,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0007015175920649888,
+      "loss": 2.7274,
+      "step": 18194
+    },
+    {
+      "epoch": 0.9278259338375843,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007008839168309675,
+      "loss": 2.7176,
+      "step": 18195
+    },
+    {
+      "epoch": 0.9278769272936896,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007002504877917995,
+      "loss": 2.7218,
+      "step": 18196
+    },
+    {
+      "epoch": 0.927927920749795,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0006996173053382149,
+      "loss": 2.7392,
+      "step": 18197
+    },
+    {
+      "epoch": 0.9279789142059004,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0006989843698607934,
+      "loss": 2.714,
+      "step": 18198
+    },
+    {
+      "epoch": 0.9280299076620058,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006983516817499607,
+      "loss": 2.753,
+      "step": 18199
+    },
+    {
+      "epoch": 0.9280809011181113,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000697719241395991,
+      "loss": 2.7294,
+      "step": 18200
+    },
+    {
+      "epoch": 0.9281318945742166,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0006970870491890052,
+      "loss": 2.7275,
+      "step": 18201
+    },
+    {
+      "epoch": 0.928182888030322,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006964551055189712,
+      "loss": 2.7366,
+      "step": 18202
+    },
+    {
+      "epoch": 0.9282338814864274,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0006958234107757033,
+      "loss": 2.7491,
+      "step": 18203
+    },
+    {
+      "epoch": 0.9282848749425328,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006951919653488631,
+      "loss": 2.7333,
+      "step": 18204
+    },
+    {
+      "epoch": 0.9283358683986381,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0006945607696279573,
+      "loss": 2.7407,
+      "step": 18205
+    },
+    {
+      "epoch": 0.9283868618547435,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0006939298240023394,
+      "loss": 2.7417,
+      "step": 18206
+    },
+    {
+      "epoch": 0.928437855310849,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0006932991288612086,
+      "loss": 2.7297,
+      "step": 18207
+    },
+    {
+      "epoch": 0.9284888487669544,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0006926686845936092,
+      "loss": 2.7001,
+      "step": 18208
+    },
+    {
+      "epoch": 0.9285398422230597,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0006920384915884306,
+      "loss": 2.7507,
+      "step": 18209
+    },
+    {
+      "epoch": 0.9285908356791651,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.000691408550234408,
+      "loss": 2.7276,
+      "step": 18210
+    },
+    {
+      "epoch": 0.9286418291352705,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0006907788609201206,
+      "loss": 2.742,
+      "step": 18211
+    },
+    {
+      "epoch": 0.9286928225913759,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0006901494240339923,
+      "loss": 2.7026,
+      "step": 18212
+    },
+    {
+      "epoch": 0.9287438160474812,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0006895202399642913,
+      "loss": 2.7347,
+      "step": 18213
+    },
+    {
+      "epoch": 0.9287948095035867,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006888913090991304,
+      "loss": 2.721,
+      "step": 18214
+    },
+    {
+      "epoch": 0.9288458029596921,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0006882626318264655,
+      "loss": 2.7329,
+      "step": 18215
+    },
+    {
+      "epoch": 0.9288967964157975,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006876342085340968,
+      "loss": 2.7785,
+      "step": 18216
+    },
+    {
+      "epoch": 0.9289477898719029,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0006870060396096662,
+      "loss": 2.7301,
+      "step": 18217
+    },
+    {
+      "epoch": 0.9289987833280082,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0006863781254406611,
+      "loss": 2.7304,
+      "step": 18218
+    },
+    {
+      "epoch": 0.9290497767841136,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0006857504664144098,
+      "loss": 2.7158,
+      "step": 18219
+    },
+    {
+      "epoch": 0.929100770240219,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.000685123062918084,
+      "loss": 2.7328,
+      "step": 18220
+    },
+    {
+      "epoch": 0.9291517636963245,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0006844959153386975,
+      "loss": 2.7335,
+      "step": 18221
+    },
+    {
+      "epoch": 0.9292027571524298,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0006838690240631076,
+      "loss": 2.7621,
+      "step": 18222
+    },
+    {
+      "epoch": 0.9292537506085352,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0006832423894780111,
+      "loss": 2.7405,
+      "step": 18223
+    },
+    {
+      "epoch": 0.9293047440646406,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000682616011969948,
+      "loss": 2.7169,
+      "step": 18224
+    },
+    {
+      "epoch": 0.929355737520746,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0006819898919252992,
+      "loss": 2.7191,
+      "step": 18225
+    },
+    {
+      "epoch": 0.9294067309768513,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006813640297302874,
+      "loss": 2.7157,
+      "step": 18226
+    },
+    {
+      "epoch": 0.9294577244329567,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006807384257709762,
+      "loss": 2.7426,
+      "step": 18227
+    },
+    {
+      "epoch": 0.9295087178890622,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0006801130804332686,
+      "loss": 2.7174,
+      "step": 18228
+    },
+    {
+      "epoch": 0.9295597113451676,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0006794879941029097,
+      "loss": 2.7112,
+      "step": 18229
+    },
+    {
+      "epoch": 0.929610704801273,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006788631671654842,
+      "loss": 2.7475,
+      "step": 18230
+    },
+    {
+      "epoch": 0.9296616982573783,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0006782386000064164,
+      "loss": 2.7389,
+      "step": 18231
+    },
+    {
+      "epoch": 0.9297126917134837,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0006776142930109705,
+      "loss": 2.7625,
+      "step": 18232
+    },
+    {
+      "epoch": 0.9297636851695891,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0006769902465642508,
+      "loss": 2.708,
+      "step": 18233
+    },
+    {
+      "epoch": 0.9298146786256944,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0006763664610512007,
+      "loss": 2.7224,
+      "step": 18234
+    },
+    {
+      "epoch": 0.9298656720817999,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0006757429368566022,
+      "loss": 2.7266,
+      "step": 18235
+    },
+    {
+      "epoch": 0.9299166655379053,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0006751196743650762,
+      "loss": 2.7389,
+      "step": 18236
+    },
+    {
+      "epoch": 0.9299676589940107,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0006744966739610826,
+      "loss": 2.7197,
+      "step": 18237
+    },
+    {
+      "epoch": 0.9300186524501161,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0006738739360289192,
+      "loss": 2.725,
+      "step": 18238
+    },
+    {
+      "epoch": 0.9300696459062214,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0006732514609527221,
+      "loss": 2.7465,
+      "step": 18239
+    },
+    {
+      "epoch": 0.9301206393623268,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0006726292491164651,
+      "loss": 2.6808,
+      "step": 18240
+    },
+    {
+      "epoch": 0.9301716328184322,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006720073009039598,
+      "loss": 2.6951,
+      "step": 18241
+    },
+    {
+      "epoch": 0.9302226262745377,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0006713856166988548,
+      "loss": 2.7126,
+      "step": 18242
+    },
+    {
+      "epoch": 0.930273619730643,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006707641968846368,
+      "loss": 2.7208,
+      "step": 18243
+    },
+    {
+      "epoch": 0.9303246131867484,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0006701430418446282,
+      "loss": 2.731,
+      "step": 18244
+    },
+    {
+      "epoch": 0.9303756066428538,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0006695221519619894,
+      "loss": 2.7174,
+      "step": 18245
+    },
+    {
+      "epoch": 0.9304266000989592,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0006689015276197152,
+      "loss": 2.7365,
+      "step": 18246
+    },
+    {
+      "epoch": 0.9304775935550645,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000668281169200639,
+      "loss": 2.7075,
+      "step": 18247
+    },
+    {
+      "epoch": 0.9305285870111699,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006676610770874285,
+      "loss": 2.7236,
+      "step": 18248
+    },
+    {
+      "epoch": 0.9305795804672754,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006670412516625878,
+      "loss": 2.7299,
+      "step": 18249
+    },
+    {
+      "epoch": 0.9306305739233808,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0006664216933084562,
+      "loss": 2.734,
+      "step": 18250
+    },
+    {
+      "epoch": 0.9306815673794862,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0006658024024072087,
+      "loss": 2.7268,
+      "step": 18251
+    },
+    {
+      "epoch": 0.9307325608355915,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0006651833793408555,
+      "loss": 2.731,
+      "step": 18252
+    },
+    {
+      "epoch": 0.9307835542916969,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0006645646244912403,
+      "loss": 2.7404,
+      "step": 18253
+    },
+    {
+      "epoch": 0.9308345477478023,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0006639461382400419,
+      "loss": 2.7217,
+      "step": 18254
+    },
+    {
+      "epoch": 0.9308855412039077,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0006633279209687746,
+      "loss": 2.7185,
+      "step": 18255
+    },
+    {
+      "epoch": 0.9309365346600131,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0006627099730587856,
+      "loss": 2.7054,
+      "step": 18256
+    },
+    {
+      "epoch": 0.9309875281161185,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0006620922948912561,
+      "loss": 2.7379,
+      "step": 18257
+    },
+    {
+      "epoch": 0.9310385215722239,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006614748868472007,
+      "loss": 2.7458,
+      "step": 18258
+    },
+    {
+      "epoch": 0.9310895150283293,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0006608577493074689,
+      "loss": 2.7045,
+      "step": 18259
+    },
+    {
+      "epoch": 0.9311405084844346,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0006602408826527408,
+      "loss": 2.7209,
+      "step": 18260
+    },
+    {
+      "epoch": 0.93119150194054,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0006596242872635313,
+      "loss": 2.7163,
+      "step": 18261
+    },
+    {
+      "epoch": 0.9312424953966454,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0006590079635201871,
+      "loss": 2.734,
+      "step": 18262
+    },
+    {
+      "epoch": 0.9312934888527509,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006583919118028882,
+      "loss": 2.7302,
+      "step": 18263
+    },
+    {
+      "epoch": 0.9313444823088562,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.000657776132491646,
+      "loss": 2.7162,
+      "step": 18264
+    },
+    {
+      "epoch": 0.9313954757649616,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000657160625966304,
+      "loss": 2.7373,
+      "step": 18265
+    },
+    {
+      "epoch": 0.931446469221067,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0006565453926065377,
+      "loss": 2.7328,
+      "step": 18266
+    },
+    {
+      "epoch": 0.9314974626771724,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0006559304327918541,
+      "loss": 2.7095,
+      "step": 18267
+    },
+    {
+      "epoch": 0.9315484561332777,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0006553157469015907,
+      "loss": 2.7328,
+      "step": 18268
+    },
+    {
+      "epoch": 0.9315994495893831,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0006547013353149173,
+      "loss": 2.7276,
+      "step": 18269
+    },
+    {
+      "epoch": 0.9316504430454886,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006540871984108331,
+      "loss": 2.7385,
+      "step": 18270
+    },
+    {
+      "epoch": 0.931701436501594,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0006534733365681695,
+      "loss": 2.7325,
+      "step": 18271
+    },
+    {
+      "epoch": 0.9317524299576994,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0006528597501655871,
+      "loss": 2.7327,
+      "step": 18272
+    },
+    {
+      "epoch": 0.9318034234138047,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0006522464395815763,
+      "loss": 2.734,
+      "step": 18273
+    },
+    {
+      "epoch": 0.9318544168699101,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006516334051944584,
+      "loss": 2.712,
+      "step": 18274
+    },
+    {
+      "epoch": 0.9319054103260155,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0006510206473823836,
+      "loss": 2.7417,
+      "step": 18275
+    },
+    {
+      "epoch": 0.9319564037821209,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006504081665233317,
+      "loss": 2.7858,
+      "step": 18276
+    },
+    {
+      "epoch": 0.9320073972382263,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0006497959629951116,
+      "loss": 2.7398,
+      "step": 18277
+    },
+    {
+      "epoch": 0.9320583906943317,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.000649184037175361,
+      "loss": 2.7208,
+      "step": 18278
+    },
+    {
+      "epoch": 0.9321093841504371,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006485723894415472,
+      "loss": 2.7178,
+      "step": 18279
+    },
+    {
+      "epoch": 0.9321603776065425,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0006479610201709647,
+      "loss": 2.7356,
+      "step": 18280
+    },
+    {
+      "epoch": 0.9322113710626478,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0006473499297407374,
+      "loss": 2.7255,
+      "step": 18281
+    },
+    {
+      "epoch": 0.9322623645187532,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006467391185278153,
+      "loss": 2.7137,
+      "step": 18282
+    },
+    {
+      "epoch": 0.9323133579748586,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0006461285869089787,
+      "loss": 2.739,
+      "step": 18283
+    },
+    {
+      "epoch": 0.9323643514309641,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0006455183352608336,
+      "loss": 2.7129,
+      "step": 18284
+    },
+    {
+      "epoch": 0.9324153448870695,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0006449083639598141,
+      "loss": 2.7258,
+      "step": 18285
+    },
+    {
+      "epoch": 0.9324663383431748,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0006442986733821805,
+      "loss": 2.7214,
+      "step": 18286
+    },
+    {
+      "epoch": 0.9325173317992802,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0006436892639040217,
+      "loss": 2.7165,
+      "step": 18287
+    },
+    {
+      "epoch": 0.9325683252553856,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0006430801359012518,
+      "loss": 2.7167,
+      "step": 18288
+    },
+    {
+      "epoch": 0.932619318711491,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0006424712897496108,
+      "loss": 2.7273,
+      "step": 18289
+    },
+    {
+      "epoch": 0.9326703121675963,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0006418627258246659,
+      "loss": 2.728,
+      "step": 18290
+    },
+    {
+      "epoch": 0.9327213056237018,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0006412544445018102,
+      "loss": 2.7137,
+      "step": 18291
+    },
+    {
+      "epoch": 0.9327722990798072,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0006406464461562622,
+      "loss": 2.7197,
+      "step": 18292
+    },
+    {
+      "epoch": 0.9328232925359126,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0006400387311630657,
+      "loss": 2.746,
+      "step": 18293
+    },
+    {
+      "epoch": 0.9328742859920179,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0006394312998970895,
+      "loss": 2.7133,
+      "step": 18294
+    },
+    {
+      "epoch": 0.9329252794481233,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006388241527330286,
+      "loss": 2.6988,
+      "step": 18295
+    },
+    {
+      "epoch": 0.9329762729042287,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000638217290045401,
+      "loss": 2.7128,
+      "step": 18296
+    },
+    {
+      "epoch": 0.9330272663603341,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006376107122085504,
+      "loss": 2.7455,
+      "step": 18297
+    },
+    {
+      "epoch": 0.9330782598164395,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006370044195966443,
+      "loss": 2.7175,
+      "step": 18298
+    },
+    {
+      "epoch": 0.9331292532725449,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0006363984125836748,
+      "loss": 2.7322,
+      "step": 18299
+    },
+    {
+      "epoch": 0.9331802467286503,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0006357926915434573,
+      "loss": 2.7401,
+      "step": 18300
+    },
+    {
+      "epoch": 0.9332312401847557,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0006351872568496311,
+      "loss": 2.7364,
+      "step": 18301
+    },
+    {
+      "epoch": 0.933282233640861,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0006345821088756587,
+      "loss": 2.7121,
+      "step": 18302
+    },
+    {
+      "epoch": 0.9333332270969664,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0006339772479948256,
+      "loss": 2.7491,
+      "step": 18303
+    },
+    {
+      "epoch": 0.9333842205530718,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0006333726745802405,
+      "loss": 2.7185,
+      "step": 18304
+    },
+    {
+      "epoch": 0.9334352140091773,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0006327683890048347,
+      "loss": 2.7308,
+      "step": 18305
+    },
+    {
+      "epoch": 0.9334862074652827,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0006321643916413614,
+      "loss": 2.7102,
+      "step": 18306
+    },
+    {
+      "epoch": 0.933537200921388,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006315606828623975,
+      "loss": 2.7247,
+      "step": 18307
+    },
+    {
+      "epoch": 0.9335881943774934,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0006309572630403401,
+      "loss": 2.7311,
+      "step": 18308
+    },
+    {
+      "epoch": 0.9336391878335988,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0006303541325474092,
+      "loss": 2.705,
+      "step": 18309
+    },
+    {
+      "epoch": 0.9336901812897042,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.000629751291755646,
+      "loss": 2.6987,
+      "step": 18310
+    },
+    {
+      "epoch": 0.9337411747458095,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.000629148741036913,
+      "loss": 2.7342,
+      "step": 18311
+    },
+    {
+      "epoch": 0.933792168201915,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006285464807628937,
+      "loss": 2.7495,
+      "step": 18312
+    },
+    {
+      "epoch": 0.9338431616580204,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0006279445113050923,
+      "loss": 2.6967,
+      "step": 18313
+    },
+    {
+      "epoch": 0.9338941551141258,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0006273428330348338,
+      "loss": 2.7371,
+      "step": 18314
+    },
+    {
+      "epoch": 0.9339451485702311,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0006267414463232639,
+      "loss": 2.7172,
+      "step": 18315
+    },
+    {
+      "epoch": 0.9339961420263365,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0006261403515413482,
+      "loss": 2.7435,
+      "step": 18316
+    },
+    {
+      "epoch": 0.9340471354824419,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0006255395490598718,
+      "loss": 2.7269,
+      "step": 18317
+    },
+    {
+      "epoch": 0.9340981289385473,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0006249390392494403,
+      "loss": 2.7065,
+      "step": 18318
+    },
+    {
+      "epoch": 0.9341491223946528,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0006243388224804779,
+      "loss": 2.7714,
+      "step": 18319
+    },
+    {
+      "epoch": 0.9342001158507581,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0006237388991232287,
+      "loss": 2.7311,
+      "step": 18320
+    },
+    {
+      "epoch": 0.9342511093068635,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0006231392695477552,
+      "loss": 2.7352,
+      "step": 18321
+    },
+    {
+      "epoch": 0.9343021027629689,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0006225399341239392,
+      "loss": 2.7602,
+      "step": 18322
+    },
+    {
+      "epoch": 0.9343530962190743,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0006219408932214813,
+      "loss": 2.7075,
+      "step": 18323
+    },
+    {
+      "epoch": 0.9344040896751796,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006213421472099002,
+      "loss": 2.756,
+      "step": 18324
+    },
+    {
+      "epoch": 0.934455083131285,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006207436964585317,
+      "loss": 2.7272,
+      "step": 18325
+    },
+    {
+      "epoch": 0.9345060765873905,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0006201455413365302,
+      "loss": 2.7369,
+      "step": 18326
+    },
+    {
+      "epoch": 0.9345570700434959,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0006195476822128687,
+      "loss": 2.7151,
+      "step": 18327
+    },
+    {
+      "epoch": 0.9346080634996012,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006189501194563365,
+      "loss": 2.7289,
+      "step": 18328
+    },
+    {
+      "epoch": 0.9346590569557066,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0006183528534355401,
+      "loss": 2.7622,
+      "step": 18329
+    },
+    {
+      "epoch": 0.934710050411812,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006177558845189029,
+      "loss": 2.7249,
+      "step": 18330
+    },
+    {
+      "epoch": 0.9347610438679174,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0006171592130746664,
+      "loss": 2.7524,
+      "step": 18331
+    },
+    {
+      "epoch": 0.9348120373240227,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006165628394708864,
+      "loss": 2.7414,
+      "step": 18332
+    },
+    {
+      "epoch": 0.9348630307801282,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0006159667640754366,
+      "loss": 2.6957,
+      "step": 18333
+    },
+    {
+      "epoch": 0.9349140242362336,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0006153709872560059,
+      "loss": 2.7194,
+      "step": 18334
+    },
+    {
+      "epoch": 0.934965017692339,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006147755093800998,
+      "loss": 2.7488,
+      "step": 18335
+    },
+    {
+      "epoch": 0.9350160111484443,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0006141803308150389,
+      "loss": 2.7077,
+      "step": 18336
+    },
+    {
+      "epoch": 0.9350670046045497,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0006135854519279592,
+      "loss": 2.7105,
+      "step": 18337
+    },
+    {
+      "epoch": 0.9351179980606551,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006129908730858118,
+      "loss": 2.7529,
+      "step": 18338
+    },
+    {
+      "epoch": 0.9351689915167605,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0006123965946553629,
+      "loss": 2.7409,
+      "step": 18339
+    },
+    {
+      "epoch": 0.935219984972866,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0006118026170031934,
+      "loss": 2.7114,
+      "step": 18340
+    },
+    {
+      "epoch": 0.9352709784289713,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0006112089404956982,
+      "loss": 2.7064,
+      "step": 18341
+    },
+    {
+      "epoch": 0.9353219718850767,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0006106155654990866,
+      "loss": 2.755,
+      "step": 18342
+    },
+    {
+      "epoch": 0.9353729653411821,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0006100224923793831,
+      "loss": 2.7003,
+      "step": 18343
+    },
+    {
+      "epoch": 0.9354239587972875,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0006094297215024243,
+      "loss": 2.7659,
+      "step": 18344
+    },
+    {
+      "epoch": 0.9354749522533928,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006088372532338611,
+      "loss": 2.7207,
+      "step": 18345
+    },
+    {
+      "epoch": 0.9355259457094982,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0006082450879391579,
+      "loss": 2.7397,
+      "step": 18346
+    },
+    {
+      "epoch": 0.9355769391656037,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0006076532259835919,
+      "loss": 2.726,
+      "step": 18347
+    },
+    {
+      "epoch": 0.9356279326217091,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0006070616677322532,
+      "loss": 2.7387,
+      "step": 18348
+    },
+    {
+      "epoch": 0.9356789260778144,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0006064704135500446,
+      "loss": 2.6954,
+      "step": 18349
+    },
+    {
+      "epoch": 0.9357299195339198,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006058794638016813,
+      "loss": 2.7281,
+      "step": 18350
+    },
+    {
+      "epoch": 0.9357809129900252,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0006052888188516914,
+      "loss": 2.7119,
+      "step": 18351
+    },
+    {
+      "epoch": 0.9358319064461306,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.0006046984790644139,
+      "loss": 2.7534,
+      "step": 18352
+    },
+    {
+      "epoch": 0.9358828999022359,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006041084448040001,
+      "loss": 2.7014,
+      "step": 18353
+    },
+    {
+      "epoch": 0.9359338933583414,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0006035187164344129,
+      "loss": 2.7089,
+      "step": 18354
+    },
+    {
+      "epoch": 0.9359848868144468,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0006029292943194262,
+      "loss": 2.7201,
+      "step": 18355
+    },
+    {
+      "epoch": 0.9360358802705522,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006023401788226253,
+      "loss": 2.716,
+      "step": 18356
+    },
+    {
+      "epoch": 0.9360868737266576,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0006017513703074061,
+      "loss": 2.7224,
+      "step": 18357
+    },
+    {
+      "epoch": 0.9361378671827629,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0006011628691369753,
+      "loss": 2.7537,
+      "step": 18358
+    },
+    {
+      "epoch": 0.9361888606388683,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006005746756743502,
+      "loss": 2.7239,
+      "step": 18359
+    },
+    {
+      "epoch": 0.9362398540949737,
+      "grad_norm": 7.96875,
+      "learning_rate": 0.0005999867902823583,
+      "loss": 2.711,
+      "step": 18360
+    },
+    {
+      "epoch": 0.9362908475510792,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0005993992133236361,
+      "loss": 2.7301,
+      "step": 18361
+    },
+    {
+      "epoch": 0.9363418410071845,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005988119451606312,
+      "loss": 2.7413,
+      "step": 18362
+    },
+    {
+      "epoch": 0.9363928344632899,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005982249861555993,
+      "loss": 2.7377,
+      "step": 18363
+    },
+    {
+      "epoch": 0.9364438279193953,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005976383366706072,
+      "loss": 2.7302,
+      "step": 18364
+    },
+    {
+      "epoch": 0.9364948213755007,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005970519970675297,
+      "loss": 2.7063,
+      "step": 18365
+    },
+    {
+      "epoch": 0.936545814831606,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005964659677080499,
+      "loss": 2.7079,
+      "step": 18366
+    },
+    {
+      "epoch": 0.9365968082877114,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005958802489536606,
+      "loss": 2.7484,
+      "step": 18367
+    },
+    {
+      "epoch": 0.9366478017438169,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0005952948411656622,
+      "loss": 2.7431,
+      "step": 18368
+    },
+    {
+      "epoch": 0.9366987951999223,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005947097447051641,
+      "loss": 2.7237,
+      "step": 18369
+    },
+    {
+      "epoch": 0.9367497886560276,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005941249599330827,
+      "loss": 2.7277,
+      "step": 18370
+    },
+    {
+      "epoch": 0.936800782112133,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005935404872101426,
+      "loss": 2.7088,
+      "step": 18371
+    },
+    {
+      "epoch": 0.9368517755682384,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005929563268968767,
+      "loss": 2.7445,
+      "step": 18372
+    },
+    {
+      "epoch": 0.9369027690243438,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005923724793536238,
+      "loss": 2.7149,
+      "step": 18373
+    },
+    {
+      "epoch": 0.9369537624804491,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005917889449405308,
+      "loss": 2.7378,
+      "step": 18374
+    },
+    {
+      "epoch": 0.9370047559365546,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0005912057240175506,
+      "loss": 2.7164,
+      "step": 18375
+    },
+    {
+      "epoch": 0.93705574939266,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005906228169444433,
+      "loss": 2.7494,
+      "step": 18376
+    },
+    {
+      "epoch": 0.9371067428487654,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005900402240807754,
+      "loss": 2.7395,
+      "step": 18377
+    },
+    {
+      "epoch": 0.9371577363048708,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0005894579457859194,
+      "loss": 2.7098,
+      "step": 18378
+    },
+    {
+      "epoch": 0.9372087297609761,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005888759824190535,
+      "loss": 2.7264,
+      "step": 18379
+    },
+    {
+      "epoch": 0.9372597232170815,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0005882943343391624,
+      "loss": 2.7177,
+      "step": 18380
+    },
+    {
+      "epoch": 0.9373107166731869,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0005877130019050358,
+      "loss": 2.7187,
+      "step": 18381
+    },
+    {
+      "epoch": 0.9373617101292924,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005871319854752686,
+      "loss": 2.7292,
+      "step": 18382
+    },
+    {
+      "epoch": 0.9374127035853977,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005865512854082611,
+      "loss": 2.7219,
+      "step": 18383
+    },
+    {
+      "epoch": 0.9374636970415031,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.000585970902062218,
+      "loss": 2.6921,
+      "step": 18384
+    },
+    {
+      "epoch": 0.9375146904976085,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005853908357951492,
+      "loss": 2.7394,
+      "step": 18385
+    },
+    {
+      "epoch": 0.9375656839537139,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0005848110869648684,
+      "loss": 2.7194,
+      "step": 18386
+    },
+    {
+      "epoch": 0.9376166774098192,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005842316559289938,
+      "loss": 2.7283,
+      "step": 18387
+    },
+    {
+      "epoch": 0.9376676708659246,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005836525430449479,
+      "loss": 2.7125,
+      "step": 18388
+    },
+    {
+      "epoch": 0.9377186643220301,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005830737486699567,
+      "loss": 2.7287,
+      "step": 18389
+    },
+    {
+      "epoch": 0.9377696577781355,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005824952731610496,
+      "loss": 2.7213,
+      "step": 18390
+    },
+    {
+      "epoch": 0.9378206512342409,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005819171168750586,
+      "loss": 2.7481,
+      "step": 18391
+    },
+    {
+      "epoch": 0.9378716446903462,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005813392801686205,
+      "loss": 2.7159,
+      "step": 18392
+    },
+    {
+      "epoch": 0.9379226381464516,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005807617633981736,
+      "loss": 2.7344,
+      "step": 18393
+    },
+    {
+      "epoch": 0.937973631602557,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005801845669199594,
+      "loss": 2.7644,
+      "step": 18394
+    },
+    {
+      "epoch": 0.9380246250586624,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0005796076910900211,
+      "loss": 2.7042,
+      "step": 18395
+    },
+    {
+      "epoch": 0.9380756185147678,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005790311362642057,
+      "loss": 2.7363,
+      "step": 18396
+    },
+    {
+      "epoch": 0.9381266119708732,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005784549027981609,
+      "loss": 2.7171,
+      "step": 18397
+    },
+    {
+      "epoch": 0.9381776054269786,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0005778789910473359,
+      "loss": 2.7293,
+      "step": 18398
+    },
+    {
+      "epoch": 0.938228598883084,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.000577303401366982,
+      "loss": 2.7109,
+      "step": 18399
+    },
+    {
+      "epoch": 0.9382795923391893,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005767281341121526,
+      "loss": 2.7262,
+      "step": 18400
+    },
+    {
+      "epoch": 0.9383305857952947,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0005761531896377009,
+      "loss": 2.7297,
+      "step": 18401
+    },
+    {
+      "epoch": 0.9383815792514001,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005755785682982817,
+      "loss": 2.6864,
+      "step": 18402
+    },
+    {
+      "epoch": 0.9384325727075056,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005750042704483503,
+      "loss": 2.7378,
+      "step": 18403
+    },
+    {
+      "epoch": 0.938483566163611,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005744302964421624,
+      "loss": 2.7129,
+      "step": 18404
+    },
+    {
+      "epoch": 0.9385345596197163,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0005738566466337741,
+      "loss": 2.7,
+      "step": 18405
+    },
+    {
+      "epoch": 0.9385855530758217,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0005732833213770413,
+      "loss": 2.7181,
+      "step": 18406
+    },
+    {
+      "epoch": 0.9386365465319271,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005727103210256195,
+      "loss": 2.711,
+      "step": 18407
+    },
+    {
+      "epoch": 0.9386875399880324,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005721376459329648,
+      "loss": 2.7124,
+      "step": 18408
+    },
+    {
+      "epoch": 0.9387385334441378,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005715652964523319,
+      "loss": 2.7054,
+      "step": 18409
+    },
+    {
+      "epoch": 0.9387895269002433,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005709932729367744,
+      "loss": 2.7178,
+      "step": 18410
+    },
+    {
+      "epoch": 0.9388405203563487,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005704215757391452,
+      "loss": 2.7323,
+      "step": 18411
+    },
+    {
+      "epoch": 0.9388915138124541,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0005698502052120957,
+      "loss": 2.7467,
+      "step": 18412
+    },
+    {
+      "epoch": 0.9389425072685594,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005692791617080764,
+      "loss": 2.7125,
+      "step": 18413
+    },
+    {
+      "epoch": 0.9389935007246648,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0005687084455793352,
+      "loss": 2.7284,
+      "step": 18414
+    },
+    {
+      "epoch": 0.9390444941807702,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005681380571779184,
+      "loss": 2.7442,
+      "step": 18415
+    },
+    {
+      "epoch": 0.9390954876368756,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005675679968556707,
+      "loss": 2.7535,
+      "step": 18416
+    },
+    {
+      "epoch": 0.939146481092981,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005669982649642337,
+      "loss": 2.7132,
+      "step": 18417
+    },
+    {
+      "epoch": 0.9391974745490864,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005664288618550468,
+      "loss": 2.7566,
+      "step": 18418
+    },
+    {
+      "epoch": 0.9392484680051918,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0005658597878793461,
+      "loss": 2.7302,
+      "step": 18419
+    },
+    {
+      "epoch": 0.9392994614612972,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005652910433881653,
+      "loss": 2.7225,
+      "step": 18420
+    },
+    {
+      "epoch": 0.9393504549174025,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005647226287323347,
+      "loss": 2.7162,
+      "step": 18421
+    },
+    {
+      "epoch": 0.9394014483735079,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0005641545442624807,
+      "loss": 2.7245,
+      "step": 18422
+    },
+    {
+      "epoch": 0.9394524418296133,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005635867903290261,
+      "loss": 2.7071,
+      "step": 18423
+    },
+    {
+      "epoch": 0.9395034352857188,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005630193672821909,
+      "loss": 2.7071,
+      "step": 18424
+    },
+    {
+      "epoch": 0.9395544287418242,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0005624522754719896,
+      "loss": 2.7466,
+      "step": 18425
+    },
+    {
+      "epoch": 0.9396054221979295,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0005618855152482334,
+      "loss": 2.7287,
+      "step": 18426
+    },
+    {
+      "epoch": 0.9396564156540349,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0005613190869605273,
+      "loss": 2.7333,
+      "step": 18427
+    },
+    {
+      "epoch": 0.9397074091101403,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005607529909582739,
+      "loss": 2.7294,
+      "step": 18428
+    },
+    {
+      "epoch": 0.9397584025662457,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005601872275906692,
+      "loss": 2.7453,
+      "step": 18429
+    },
+    {
+      "epoch": 0.939809396022351,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0005596217972067046,
+      "loss": 2.7385,
+      "step": 18430
+    },
+    {
+      "epoch": 0.9398603894784565,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005590567001551654,
+      "loss": 2.7207,
+      "step": 18431
+    },
+    {
+      "epoch": 0.9399113829345619,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0005584919367846329,
+      "loss": 2.717,
+      "step": 18432
+    },
+    {
+      "epoch": 0.9399623763906673,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0005579275074434811,
+      "loss": 2.7355,
+      "step": 18433
+    },
+    {
+      "epoch": 0.9400133698467726,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005573634124798781,
+      "loss": 2.7301,
+      "step": 18434
+    },
+    {
+      "epoch": 0.940064363302878,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0005567996522417857,
+      "loss": 2.7277,
+      "step": 18435
+    },
+    {
+      "epoch": 0.9401153567589834,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005562362270769604,
+      "loss": 2.7228,
+      "step": 18436
+    },
+    {
+      "epoch": 0.9401663502150888,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0005556731373329508,
+      "loss": 2.7329,
+      "step": 18437
+    },
+    {
+      "epoch": 0.9402173436711943,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005551103833570988,
+      "loss": 2.7553,
+      "step": 18438
+    },
+    {
+      "epoch": 0.9402683371272996,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005545479654965393,
+      "loss": 2.7034,
+      "step": 18439
+    },
+    {
+      "epoch": 0.940319330583405,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005539858840982004,
+      "loss": 2.7231,
+      "step": 18440
+    },
+    {
+      "epoch": 0.9403703240395104,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005534241395088017,
+      "loss": 2.7269,
+      "step": 18441
+    },
+    {
+      "epoch": 0.9404213174956157,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005528627320748554,
+      "loss": 2.7254,
+      "step": 18442
+    },
+    {
+      "epoch": 0.9404723109517211,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005523016621426657,
+      "loss": 2.7288,
+      "step": 18443
+    },
+    {
+      "epoch": 0.9405233044078265,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0005517409300583294,
+      "loss": 2.7315,
+      "step": 18444
+    },
+    {
+      "epoch": 0.940574297863932,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005511805361677335,
+      "loss": 2.7034,
+      "step": 18445
+    },
+    {
+      "epoch": 0.9406252913200374,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005506204808165578,
+      "loss": 2.7072,
+      "step": 18446
+    },
+    {
+      "epoch": 0.9406762847761427,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0005500607643502721,
+      "loss": 2.7444,
+      "step": 18447
+    },
+    {
+      "epoch": 0.9407272782322481,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005495013871141377,
+      "loss": 2.7225,
+      "step": 18448
+    },
+    {
+      "epoch": 0.9407782716883535,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0005489423494532067,
+      "loss": 2.73,
+      "step": 18449
+    },
+    {
+      "epoch": 0.9408292651444589,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005483836517123214,
+      "loss": 2.719,
+      "step": 18450
+    },
+    {
+      "epoch": 0.9408802586005642,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005478252942361145,
+      "loss": 2.7233,
+      "step": 18451
+    },
+    {
+      "epoch": 0.9409312520566697,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0005472672773690096,
+      "loss": 2.7226,
+      "step": 18452
+    },
+    {
+      "epoch": 0.9409822455127751,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005467096014552191,
+      "loss": 2.7446,
+      "step": 18453
+    },
+    {
+      "epoch": 0.9410332389688805,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005461522668387456,
+      "loss": 2.7341,
+      "step": 18454
+    },
+    {
+      "epoch": 0.9410842324249858,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005455952738633813,
+      "loss": 2.7196,
+      "step": 18455
+    },
+    {
+      "epoch": 0.9411352258810912,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0005450386228727069,
+      "loss": 2.7618,
+      "step": 18456
+    },
+    {
+      "epoch": 0.9411862193371966,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005444823142100932,
+      "loss": 2.7327,
+      "step": 18457
+    },
+    {
+      "epoch": 0.941237212793302,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005439263482186992,
+      "loss": 2.7398,
+      "step": 18458
+    },
+    {
+      "epoch": 0.9412882062494075,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005433707252414724,
+      "loss": 2.7147,
+      "step": 18459
+    },
+    {
+      "epoch": 0.9413391997055128,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005428154456211493,
+      "loss": 2.7352,
+      "step": 18460
+    },
+    {
+      "epoch": 0.9413901931616182,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005422605097002545,
+      "loss": 2.7333,
+      "step": 18461
+    },
+    {
+      "epoch": 0.9414411866177236,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005417059178210999,
+      "loss": 2.72,
+      "step": 18462
+    },
+    {
+      "epoch": 0.941492180073829,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005411516703257852,
+      "loss": 2.7287,
+      "step": 18463
+    },
+    {
+      "epoch": 0.9415431735299343,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0005405977675561992,
+      "loss": 2.7226,
+      "step": 18464
+    },
+    {
+      "epoch": 0.9415941669860397,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0005400442098540164,
+      "loss": 2.7189,
+      "step": 18465
+    },
+    {
+      "epoch": 0.9416451604421452,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000539490997560699,
+      "loss": 2.7539,
+      "step": 18466
+    },
+    {
+      "epoch": 0.9416961538982506,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005389381310174957,
+      "loss": 2.7111,
+      "step": 18467
+    },
+    {
+      "epoch": 0.9417471473543559,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005383856105654432,
+      "loss": 2.7114,
+      "step": 18468
+    },
+    {
+      "epoch": 0.9417981408104613,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0005378334365453639,
+      "loss": 2.7113,
+      "step": 18469
+    },
+    {
+      "epoch": 0.9418491342665667,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005372816092978657,
+      "loss": 2.738,
+      "step": 18470
+    },
+    {
+      "epoch": 0.9419001277226721,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005367301291633435,
+      "loss": 2.7217,
+      "step": 18471
+    },
+    {
+      "epoch": 0.9419511211787774,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005361789964819783,
+      "loss": 2.7602,
+      "step": 18472
+    },
+    {
+      "epoch": 0.9420021146348829,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005356282115937366,
+      "loss": 2.7342,
+      "step": 18473
+    },
+    {
+      "epoch": 0.9420531080909883,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0005350777748383697,
+      "loss": 2.7508,
+      "step": 18474
+    },
+    {
+      "epoch": 0.9421041015470937,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005345276865554148,
+      "loss": 2.7136,
+      "step": 18475
+    },
+    {
+      "epoch": 0.942155095003199,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0005339779470841943,
+      "loss": 2.7243,
+      "step": 18476
+    },
+    {
+      "epoch": 0.9422060884593044,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0005334285567638148,
+      "loss": 2.719,
+      "step": 18477
+    },
+    {
+      "epoch": 0.9422570819154098,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000532879515933168,
+      "loss": 2.7033,
+      "step": 18478
+    },
+    {
+      "epoch": 0.9423080753715152,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005323308249309296,
+      "loss": 2.7044,
+      "step": 18479
+    },
+    {
+      "epoch": 0.9423590688276207,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00053178248409556,
+      "loss": 2.7348,
+      "step": 18480
+    },
+    {
+      "epoch": 0.942410062283726,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0005312344937653035,
+      "loss": 2.7254,
+      "step": 18481
+    },
+    {
+      "epoch": 0.9424610557398314,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.000530686854278188,
+      "loss": 2.7155,
+      "step": 18482
+    },
+    {
+      "epoch": 0.9425120491959368,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.000530139565972025,
+      "loss": 2.7365,
+      "step": 18483
+    },
+    {
+      "epoch": 0.9425630426520422,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0005295926291844093,
+      "loss": 2.7299,
+      "step": 18484
+    },
+    {
+      "epoch": 0.9426140361081475,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0005290460442527191,
+      "loss": 2.7226,
+      "step": 18485
+    },
+    {
+      "epoch": 0.9426650295642529,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0005284998115141157,
+      "loss": 2.7006,
+      "step": 18486
+    },
+    {
+      "epoch": 0.9427160230203584,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005279539313055424,
+      "loss": 2.7047,
+      "step": 18487
+    },
+    {
+      "epoch": 0.9427670164764638,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005274084039637255,
+      "loss": 2.7186,
+      "step": 18488
+    },
+    {
+      "epoch": 0.9428180099325691,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005268632298251743,
+      "loss": 2.7255,
+      "step": 18489
+    },
+    {
+      "epoch": 0.9428690033886745,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0005263184092261792,
+      "loss": 2.6971,
+      "step": 18490
+    },
+    {
+      "epoch": 0.9429199968447799,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0005257739425028132,
+      "loss": 2.7054,
+      "step": 18491
+    },
+    {
+      "epoch": 0.9429709903008853,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005252298299909303,
+      "loss": 2.7245,
+      "step": 18492
+    },
+    {
+      "epoch": 0.9430219837569906,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005246860720261666,
+      "loss": 2.7369,
+      "step": 18493
+    },
+    {
+      "epoch": 0.9430729772130961,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0005241426689439395,
+      "loss": 2.7052,
+      "step": 18494
+    },
+    {
+      "epoch": 0.9431239706692015,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005235996210794472,
+      "loss": 2.7255,
+      "step": 18495
+    },
+    {
+      "epoch": 0.9431749641253069,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0005230569287676685,
+      "loss": 2.7069,
+      "step": 18496
+    },
+    {
+      "epoch": 0.9432259575814123,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0005225145923433643,
+      "loss": 2.7354,
+      "step": 18497
+    },
+    {
+      "epoch": 0.9432769510375176,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.000521972612141074,
+      "loss": 2.728,
+      "step": 18498
+    },
+    {
+      "epoch": 0.943327944493623,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000521430988495119,
+      "loss": 2.7096,
+      "step": 18499
+    },
+    {
+      "epoch": 0.9433789379497284,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000520889721739599,
+      "loss": 2.7388,
+      "step": 18500
+    },
+    {
+      "epoch": 0.9434299314058339,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005203488122083953,
+      "loss": 2.7095,
+      "step": 18501
+    },
+    {
+      "epoch": 0.9434809248619392,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0005198082602351679,
+      "loss": 2.7006,
+      "step": 18502
+    },
+    {
+      "epoch": 0.9435319183180446,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0005192680661533564,
+      "loss": 2.7061,
+      "step": 18503
+    },
+    {
+      "epoch": 0.94358291177415,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005187282302961794,
+      "loss": 2.7341,
+      "step": 18504
+    },
+    {
+      "epoch": 0.9436339052302554,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005181887529966353,
+      "loss": 2.7175,
+      "step": 18505
+    },
+    {
+      "epoch": 0.9436848986863607,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005176496345875004,
+      "loss": 2.7107,
+      "step": 18506
+    },
+    {
+      "epoch": 0.9437358921424661,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005171108754013301,
+      "loss": 2.7052,
+      "step": 18507
+    },
+    {
+      "epoch": 0.9437868855985716,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0005165724757704578,
+      "loss": 2.7316,
+      "step": 18508
+    },
+    {
+      "epoch": 0.943837879054677,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005160344360269962,
+      "loss": 2.7167,
+      "step": 18509
+    },
+    {
+      "epoch": 0.9438888725107824,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005154967565028348,
+      "loss": 2.7201,
+      "step": 18510
+    },
+    {
+      "epoch": 0.9439398659668877,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0005149594375296414,
+      "loss": 2.7619,
+      "step": 18511
+    },
+    {
+      "epoch": 0.9439908594229931,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0005144224794388613,
+      "loss": 2.7225,
+      "step": 18512
+    },
+    {
+      "epoch": 0.9440418528790985,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005138858825617173,
+      "loss": 2.726,
+      "step": 18513
+    },
+    {
+      "epoch": 0.9440928463352039,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0005133496472292093,
+      "loss": 2.7359,
+      "step": 18514
+    },
+    {
+      "epoch": 0.9441438397913093,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005128137737721142,
+      "loss": 2.7325,
+      "step": 18515
+    },
+    {
+      "epoch": 0.9441948332474147,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0005122782625209853,
+      "loss": 2.7257,
+      "step": 18516
+    },
+    {
+      "epoch": 0.9442458267035201,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005117431138061535,
+      "loss": 2.6936,
+      "step": 18517
+    },
+    {
+      "epoch": 0.9442968201596255,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0005112083279577251,
+      "loss": 2.7023,
+      "step": 18518
+    },
+    {
+      "epoch": 0.9443478136157308,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005106739053055829,
+      "loss": 2.7577,
+      "step": 18519
+    },
+    {
+      "epoch": 0.9443988070718362,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005101398461793856,
+      "loss": 2.6926,
+      "step": 18520
+    },
+    {
+      "epoch": 0.9444498005279416,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005096061509085678,
+      "loss": 2.6744,
+      "step": 18521
+    },
+    {
+      "epoch": 0.9445007939840471,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005090728198223393,
+      "loss": 2.7336,
+      "step": 18522
+    },
+    {
+      "epoch": 0.9445517874401524,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0005085398532496859,
+      "loss": 2.7355,
+      "step": 18523
+    },
+    {
+      "epoch": 0.9446027808962578,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005080072515193677,
+      "loss": 2.7544,
+      "step": 18524
+    },
+    {
+      "epoch": 0.9446537743523632,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005074750149599205,
+      "loss": 2.7227,
+      "step": 18525
+    },
+    {
+      "epoch": 0.9447047678084686,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0005069431438996548,
+      "loss": 2.7329,
+      "step": 18526
+    },
+    {
+      "epoch": 0.944755761264574,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.000506411638666655,
+      "loss": 2.7231,
+      "step": 18527
+    },
+    {
+      "epoch": 0.9448067547206793,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005058804995887804,
+      "loss": 2.717,
+      "step": 18528
+    },
+    {
+      "epoch": 0.9448577481767848,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0005053497269936642,
+      "loss": 2.7348,
+      "step": 18529
+    },
+    {
+      "epoch": 0.9449087416328902,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0005048193212087136,
+      "loss": 2.7236,
+      "step": 18530
+    },
+    {
+      "epoch": 0.9449597350889956,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005042892825611094,
+      "loss": 2.7258,
+      "step": 18531
+    },
+    {
+      "epoch": 0.9450107285451009,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000503759611377806,
+      "loss": 2.7376,
+      "step": 18532
+    },
+    {
+      "epoch": 0.9450617220012063,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0005032303079855318,
+      "loss": 2.7403,
+      "step": 18533
+    },
+    {
+      "epoch": 0.9451127154573117,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005027013727107873,
+      "loss": 2.7036,
+      "step": 18534
+    },
+    {
+      "epoch": 0.9451637089134171,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0005021728058798467,
+      "loss": 2.7233,
+      "step": 18535
+    },
+    {
+      "epoch": 0.9452147023695225,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005016446078187558,
+      "loss": 2.7073,
+      "step": 18536
+    },
+    {
+      "epoch": 0.9452656958256279,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005011167788533344,
+      "loss": 2.7218,
+      "step": 18537
+    },
+    {
+      "epoch": 0.9453166892817333,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0005005893193091736,
+      "loss": 2.7261,
+      "step": 18538
+    },
+    {
+      "epoch": 0.9453676827378387,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005000622295116371,
+      "loss": 2.7201,
+      "step": 18539
+    },
+    {
+      "epoch": 0.945418676193944,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0004995355097858602,
+      "loss": 2.7429,
+      "step": 18540
+    },
+    {
+      "epoch": 0.9454696696500494,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0004990091604567507,
+      "loss": 2.739,
+      "step": 18541
+    },
+    {
+      "epoch": 0.9455206631061548,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0004984831818489865,
+      "loss": 2.7407,
+      "step": 18542
+    },
+    {
+      "epoch": 0.9455716565622603,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0004979575742870179,
+      "loss": 2.712,
+      "step": 18543
+    },
+    {
+      "epoch": 0.9456226500183657,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000497432338095066,
+      "loss": 2.7234,
+      "step": 18544
+    },
+    {
+      "epoch": 0.945673643474471,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0004969074735971232,
+      "loss": 2.7168,
+      "step": 18545
+    },
+    {
+      "epoch": 0.9457246369305764,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000496382981116952,
+      "loss": 2.7303,
+      "step": 18546
+    },
+    {
+      "epoch": 0.9457756303866818,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0004958588609780859,
+      "loss": 2.7236,
+      "step": 18547
+    },
+    {
+      "epoch": 0.9458266238427872,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0004953351135038285,
+      "loss": 2.7438,
+      "step": 18548
+    },
+    {
+      "epoch": 0.9458776172988925,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0004948117390172533,
+      "loss": 2.7367,
+      "step": 18549
+    },
+    {
+      "epoch": 0.945928610754998,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0004942887378412042,
+      "loss": 2.7424,
+      "step": 18550
+    },
+    {
+      "epoch": 0.9459796042111034,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0004937661102982946,
+      "loss": 2.7125,
+      "step": 18551
+    },
+    {
+      "epoch": 0.9460305976672088,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0004932438567109071,
+      "loss": 2.6862,
+      "step": 18552
+    },
+    {
+      "epoch": 0.9460815911233141,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0004927219774011946,
+      "loss": 2.7066,
+      "step": 18553
+    },
+    {
+      "epoch": 0.9461325845794195,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0004922004726910779,
+      "loss": 2.7222,
+      "step": 18554
+    },
+    {
+      "epoch": 0.9461835780355249,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0004916793429022479,
+      "loss": 2.7091,
+      "step": 18555
+    },
+    {
+      "epoch": 0.9462345714916303,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000491158588356163,
+      "loss": 2.7228,
+      "step": 18556
+    },
+    {
+      "epoch": 0.9462855649477357,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0004906382093740511,
+      "loss": 2.7306,
+      "step": 18557
+    },
+    {
+      "epoch": 0.9463365584038411,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0004901182062769081,
+      "loss": 2.7245,
+      "step": 18558
+    },
+    {
+      "epoch": 0.9463875518599465,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000489598579385498,
+      "loss": 2.7311,
+      "step": 18559
+    },
+    {
+      "epoch": 0.9464385453160519,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004890793290203524,
+      "loss": 2.6996,
+      "step": 18560
+    },
+    {
+      "epoch": 0.9464895387721572,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004885604555017717,
+      "loss": 2.7172,
+      "step": 18561
+    },
+    {
+      "epoch": 0.9465405322282626,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00048804195914982276,
+      "loss": 2.7113,
+      "step": 18562
+    },
+    {
+      "epoch": 0.946591525684368,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0004875238402843403,
+      "loss": 2.7116,
+      "step": 18563
+    },
+    {
+      "epoch": 0.9466425191404735,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00048700609922492597,
+      "loss": 2.7482,
+      "step": 18564
+    },
+    {
+      "epoch": 0.9466935125965789,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0004864887362909485,
+      "loss": 2.7597,
+      "step": 18565
+    },
+    {
+      "epoch": 0.9467445060526842,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0004859717518015433,
+      "loss": 2.7339,
+      "step": 18566
+    },
+    {
+      "epoch": 0.9467954995087896,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0004854551460756125,
+      "loss": 2.7098,
+      "step": 18567
+    },
+    {
+      "epoch": 0.946846492964895,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00048493891943182416,
+      "loss": 2.7352,
+      "step": 18568
+    },
+    {
+      "epoch": 0.9468974864210004,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0004844230721886133,
+      "loss": 2.7309,
+      "step": 18569
+    },
+    {
+      "epoch": 0.9469484798771057,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00048390760466418016,
+      "loss": 2.7065,
+      "step": 18570
+    },
+    {
+      "epoch": 0.9469994733332112,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00048339251717649127,
+      "loss": 2.7074,
+      "step": 18571
+    },
+    {
+      "epoch": 0.9470504667893166,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0004828778100432777,
+      "loss": 2.7486,
+      "step": 18572
+    },
+    {
+      "epoch": 0.947101460245422,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00048236348358203745,
+      "loss": 2.7251,
+      "step": 18573
+    },
+    {
+      "epoch": 0.9471524537015273,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0004818495381100327,
+      "loss": 2.7167,
+      "step": 18574
+    },
+    {
+      "epoch": 0.9472034471576327,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00048133597394429067,
+      "loss": 2.712,
+      "step": 18575
+    },
+    {
+      "epoch": 0.9472544406137381,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004808227914016036,
+      "loss": 2.7548,
+      "step": 18576
+    },
+    {
+      "epoch": 0.9473054340698435,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0004803099907985285,
+      "loss": 2.7117,
+      "step": 18577
+    },
+    {
+      "epoch": 0.947356427525949,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0004797975724513866,
+      "loss": 2.6839,
+      "step": 18578
+    },
+    {
+      "epoch": 0.9474074209820543,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0004792855366762629,
+      "loss": 2.7253,
+      "step": 18579
+    },
+    {
+      "epoch": 0.9474584144381597,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00047877388378900654,
+      "loss": 2.7357,
+      "step": 18580
+    },
+    {
+      "epoch": 0.9475094078942651,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00047826261410523153,
+      "loss": 2.7066,
+      "step": 18581
+    },
+    {
+      "epoch": 0.9475604013503705,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0004777517279403143,
+      "loss": 2.7257,
+      "step": 18582
+    },
+    {
+      "epoch": 0.9476113948064758,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0004772412256093951,
+      "loss": 2.71,
+      "step": 18583
+    },
+    {
+      "epoch": 0.9476623882625812,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0004767311074273775,
+      "loss": 2.6807,
+      "step": 18584
+    },
+    {
+      "epoch": 0.9477133817186867,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004762213737089279,
+      "loss": 2.7428,
+      "step": 18585
+    },
+    {
+      "epoch": 0.9477643751747921,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00047571202476847567,
+      "loss": 2.7264,
+      "step": 18586
+    },
+    {
+      "epoch": 0.9478153686308974,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00047520306092021304,
+      "loss": 2.7429,
+      "step": 18587
+    },
+    {
+      "epoch": 0.9478663620870028,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00047469448247809394,
+      "loss": 2.7587,
+      "step": 18588
+    },
+    {
+      "epoch": 0.9479173555431082,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00047418628975583584,
+      "loss": 2.7171,
+      "step": 18589
+    },
+    {
+      "epoch": 0.9479683489992136,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00047367848306691714,
+      "loss": 2.7148,
+      "step": 18590
+    },
+    {
+      "epoch": 0.9480193424553189,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0004731710627245786,
+      "loss": 2.7113,
+      "step": 18591
+    },
+    {
+      "epoch": 0.9480703359114244,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0004726640290418226,
+      "loss": 2.7027,
+      "step": 18592
+    },
+    {
+      "epoch": 0.9481213293675298,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004721573823314129,
+      "loss": 2.7192,
+      "step": 18593
+    },
+    {
+      "epoch": 0.9481723228236352,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00047165112290587473,
+      "loss": 2.7024,
+      "step": 18594
+    },
+    {
+      "epoch": 0.9482233162797405,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0004711452510774944,
+      "loss": 2.7152,
+      "step": 18595
+    },
+    {
+      "epoch": 0.9482743097358459,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0004706397671583185,
+      "loss": 2.682,
+      "step": 18596
+    },
+    {
+      "epoch": 0.9483253031919513,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0004701346714601558,
+      "loss": 2.7513,
+      "step": 18597
+    },
+    {
+      "epoch": 0.9483762966480567,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0004696299642945743,
+      "loss": 2.6958,
+      "step": 18598
+    },
+    {
+      "epoch": 0.9484272901041622,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0004691256459729027,
+      "loss": 2.7188,
+      "step": 18599
+    },
+    {
+      "epoch": 0.9484782835602675,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0004686217168062299,
+      "loss": 2.7071,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9485292770163729,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0004681181771054048,
+      "loss": 2.727,
+      "step": 18601
+    },
+    {
+      "epoch": 0.9485802704724783,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.00046761502718103585,
+      "loss": 2.7272,
+      "step": 18602
+    },
+    {
+      "epoch": 0.9486312639285837,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00046711226734349125,
+      "loss": 2.7004,
+      "step": 18603
+    },
+    {
+      "epoch": 0.948682257384689,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00046660989790289835,
+      "loss": 2.7375,
+      "step": 18604
+    },
+    {
+      "epoch": 0.9487332508407944,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00046610791916914434,
+      "loss": 2.7133,
+      "step": 18605
+    },
+    {
+      "epoch": 0.9487842442968999,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004656063314518747,
+      "loss": 2.7238,
+      "step": 18606
+    },
+    {
+      "epoch": 0.9488352377530053,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0004651051350604942,
+      "loss": 2.7321,
+      "step": 18607
+    },
+    {
+      "epoch": 0.9488862312091106,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00046460433030416496,
+      "loss": 2.6959,
+      "step": 18608
+    },
+    {
+      "epoch": 0.948937224665216,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0004641039174918095,
+      "loss": 2.6881,
+      "step": 18609
+    },
+    {
+      "epoch": 0.9489882181213214,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0004636038969321073,
+      "loss": 2.7426,
+      "step": 18610
+    },
+    {
+      "epoch": 0.9490392115774268,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00046310426893349623,
+      "loss": 2.7281,
+      "step": 18611
+    },
+    {
+      "epoch": 0.9490902050335321,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00046260503380417175,
+      "loss": 2.7246,
+      "step": 18612
+    },
+    {
+      "epoch": 0.9491411984896376,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000462106191852087,
+      "loss": 2.7157,
+      "step": 18613
+    },
+    {
+      "epoch": 0.949192191945743,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00046160774338495375,
+      "loss": 2.7146,
+      "step": 18614
+    },
+    {
+      "epoch": 0.9492431854018484,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00046110968871023887,
+      "loss": 2.7088,
+      "step": 18615
+    },
+    {
+      "epoch": 0.9492941788579538,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00046061202813516824,
+      "loss": 2.7151,
+      "step": 18616
+    },
+    {
+      "epoch": 0.9493451723140591,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0004601147619667232,
+      "loss": 2.7266,
+      "step": 18617
+    },
+    {
+      "epoch": 0.9493961657701645,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0004596178905116432,
+      "loss": 2.7178,
+      "step": 18618
+    },
+    {
+      "epoch": 0.9494471592262699,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0004591214140764234,
+      "loss": 2.6969,
+      "step": 18619
+    },
+    {
+      "epoch": 0.9494981526823753,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0004586253329673152,
+      "loss": 2.7232,
+      "step": 18620
+    },
+    {
+      "epoch": 0.9495491461384807,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0004581296474903266,
+      "loss": 2.7026,
+      "step": 18621
+    },
+    {
+      "epoch": 0.9496001395945861,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0004576343579512209,
+      "loss": 2.6815,
+      "step": 18622
+    },
+    {
+      "epoch": 0.9496511330506915,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000457139464655518,
+      "loss": 2.7289,
+      "step": 18623
+    },
+    {
+      "epoch": 0.9497021265067969,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00045664496790849275,
+      "loss": 2.7091,
+      "step": 18624
+    },
+    {
+      "epoch": 0.9497531199629022,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0004561508680151754,
+      "loss": 2.7385,
+      "step": 18625
+    },
+    {
+      "epoch": 0.9498041134190076,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00045565716528035195,
+      "loss": 2.7336,
+      "step": 18626
+    },
+    {
+      "epoch": 0.949855106875113,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00045516386000856297,
+      "loss": 2.6825,
+      "step": 18627
+    },
+    {
+      "epoch": 0.9499061003312185,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00045467095250410387,
+      "loss": 2.7359,
+      "step": 18628
+    },
+    {
+      "epoch": 0.9499570937873238,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0004541784430710248,
+      "loss": 2.7419,
+      "step": 18629
+    },
+    {
+      "epoch": 0.9500080872434292,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0004536863320131302,
+      "loss": 2.7087,
+      "step": 18630
+    },
+    {
+      "epoch": 0.9500590806995346,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000453194619633979,
+      "loss": 2.7202,
+      "step": 18631
+    },
+    {
+      "epoch": 0.95011007415564,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0004527033062368841,
+      "loss": 2.7335,
+      "step": 18632
+    },
+    {
+      "epoch": 0.9501610676117453,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0004522123921249121,
+      "loss": 2.7021,
+      "step": 18633
+    },
+    {
+      "epoch": 0.9502120610678507,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004517218776008838,
+      "loss": 2.7132,
+      "step": 18634
+    },
+    {
+      "epoch": 0.9502630545239562,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0004512317629673731,
+      "loss": 2.7206,
+      "step": 18635
+    },
+    {
+      "epoch": 0.9503140479800616,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00045074204852670735,
+      "loss": 2.6999,
+      "step": 18636
+    },
+    {
+      "epoch": 0.950365041436167,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00045025273458096685,
+      "loss": 2.7105,
+      "step": 18637
+    },
+    {
+      "epoch": 0.9504160348922723,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00044976382143198525,
+      "loss": 2.7316,
+      "step": 18638
+    },
+    {
+      "epoch": 0.9504670283483777,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0004492753093813486,
+      "loss": 2.7276,
+      "step": 18639
+    },
+    {
+      "epoch": 0.9505180218044831,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00044878719873039577,
+      "loss": 2.726,
+      "step": 18640
+    },
+    {
+      "epoch": 0.9505690152605885,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00044829948978021756,
+      "loss": 2.7233,
+      "step": 18641
+    },
+    {
+      "epoch": 0.9506200087166939,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00044781218283165803,
+      "loss": 2.7288,
+      "step": 18642
+    },
+    {
+      "epoch": 0.9506710021727993,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0004473252781853124,
+      "loss": 2.7208,
+      "step": 18643
+    },
+    {
+      "epoch": 0.9507219956289047,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0004468387761415276,
+      "loss": 2.7438,
+      "step": 18644
+    },
+    {
+      "epoch": 0.9507729890850101,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00044635267700040226,
+      "loss": 2.7054,
+      "step": 18645
+    },
+    {
+      "epoch": 0.9508239825411154,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0004458669810617876,
+      "loss": 2.7054,
+      "step": 18646
+    },
+    {
+      "epoch": 0.9508749759972208,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00044538168862528473,
+      "loss": 2.7202,
+      "step": 18647
+    },
+    {
+      "epoch": 0.9509259694533262,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00044489679999024665,
+      "loss": 2.7184,
+      "step": 18648
+    },
+    {
+      "epoch": 0.9509769629094317,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0004444123154557766,
+      "loss": 2.7072,
+      "step": 18649
+    },
+    {
+      "epoch": 0.951027956365537,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00044392823532072976,
+      "loss": 2.721,
+      "step": 18650
+    },
+    {
+      "epoch": 0.9510789498216424,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00044344455988371055,
+      "loss": 2.7132,
+      "step": 18651
+    },
+    {
+      "epoch": 0.9511299432777478,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00044296128944307423,
+      "loss": 2.7042,
+      "step": 18652
+    },
+    {
+      "epoch": 0.9511809367338532,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0004424784242969261,
+      "loss": 2.73,
+      "step": 18653
+    },
+    {
+      "epoch": 0.9512319301899586,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0004419959647431223,
+      "loss": 2.7257,
+      "step": 18654
+    },
+    {
+      "epoch": 0.9512829236460639,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00044151391107926784,
+      "loss": 2.7413,
+      "step": 18655
+    },
+    {
+      "epoch": 0.9513339171021694,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0004410322636027175,
+      "loss": 2.7128,
+      "step": 18656
+    },
+    {
+      "epoch": 0.9513849105582748,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00044055102261057565,
+      "loss": 2.6917,
+      "step": 18657
+    },
+    {
+      "epoch": 0.9514359040143802,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00044007018839969616,
+      "loss": 2.7186,
+      "step": 18658
+    },
+    {
+      "epoch": 0.9514868974704855,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00043958976126668147,
+      "loss": 2.734,
+      "step": 18659
+    },
+    {
+      "epoch": 0.9515378909265909,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00043910974150788317,
+      "loss": 2.7016,
+      "step": 18660
+    },
+    {
+      "epoch": 0.9515888843826963,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0004386301294194016,
+      "loss": 2.7193,
+      "step": 18661
+    },
+    {
+      "epoch": 0.9516398778388017,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00043815092529708564,
+      "loss": 2.7302,
+      "step": 18662
+    },
+    {
+      "epoch": 0.9516908712949071,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0004376721294365327,
+      "loss": 2.7199,
+      "step": 18663
+    },
+    {
+      "epoch": 0.9517418647510125,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00043719374213308784,
+      "loss": 2.7404,
+      "step": 18664
+    },
+    {
+      "epoch": 0.9517928582071179,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0004367157636818444,
+      "loss": 2.7403,
+      "step": 18665
+    },
+    {
+      "epoch": 0.9518438516632233,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00043623819437764357,
+      "loss": 2.7195,
+      "step": 18666
+    },
+    {
+      "epoch": 0.9518948451193286,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0004357610345150742,
+      "loss": 2.7096,
+      "step": 18667
+    },
+    {
+      "epoch": 0.951945838575434,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0004352842843884721,
+      "loss": 2.7464,
+      "step": 18668
+    },
+    {
+      "epoch": 0.9519968320315394,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00043480794429192086,
+      "loss": 2.7223,
+      "step": 18669
+    },
+    {
+      "epoch": 0.9520478254876449,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0004343320145192512,
+      "loss": 2.7278,
+      "step": 18670
+    },
+    {
+      "epoch": 0.9520988189437503,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00043385649536404046,
+      "loss": 2.7396,
+      "step": 18671
+    },
+    {
+      "epoch": 0.9521498123998556,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0004333813871196126,
+      "loss": 2.734,
+      "step": 18672
+    },
+    {
+      "epoch": 0.952200805855961,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0004329066900790385,
+      "loss": 2.73,
+      "step": 18673
+    },
+    {
+      "epoch": 0.9522517993120664,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004324324045351349,
+      "loss": 2.7003,
+      "step": 18674
+    },
+    {
+      "epoch": 0.9523027927681718,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0004319585307804648,
+      "loss": 2.7476,
+      "step": 18675
+    },
+    {
+      "epoch": 0.9523537862242771,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00043148506910733775,
+      "loss": 2.7156,
+      "step": 18676
+    },
+    {
+      "epoch": 0.9524047796803826,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00043101201980780827,
+      "loss": 2.7271,
+      "step": 18677
+    },
+    {
+      "epoch": 0.952455773136488,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0004305393831736775,
+      "loss": 2.7138,
+      "step": 18678
+    },
+    {
+      "epoch": 0.9525067665925934,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000430067159496491,
+      "loss": 2.7512,
+      "step": 18679
+    },
+    {
+      "epoch": 0.9525577600486987,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0004295953490675404,
+      "loss": 2.7267,
+      "step": 18680
+    },
+    {
+      "epoch": 0.9526087535048041,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00042912395217786146,
+      "loss": 2.7069,
+      "step": 18681
+    },
+    {
+      "epoch": 0.9526597469609095,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00042865296911823614,
+      "loss": 2.7434,
+      "step": 18682
+    },
+    {
+      "epoch": 0.9527107404170149,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0004281824001791902,
+      "loss": 2.72,
+      "step": 18683
+    },
+    {
+      "epoch": 0.9527617338731204,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00042771224565099413,
+      "loss": 2.723,
+      "step": 18684
+    },
+    {
+      "epoch": 0.9528127273292257,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004272425058236627,
+      "loss": 2.7194,
+      "step": 18685
+    },
+    {
+      "epoch": 0.9528637207853311,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0004267731809869559,
+      "loss": 2.7552,
+      "step": 18686
+    },
+    {
+      "epoch": 0.9529147142414365,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00042630427143037597,
+      "loss": 2.7178,
+      "step": 18687
+    },
+    {
+      "epoch": 0.9529657076975419,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00042583577744317,
+      "loss": 2.7017,
+      "step": 18688
+    },
+    {
+      "epoch": 0.9530167011536472,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0004253676993143287,
+      "loss": 2.7245,
+      "step": 18689
+    },
+    {
+      "epoch": 0.9530676946097526,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00042490003733258643,
+      "loss": 2.7358,
+      "step": 18690
+    },
+    {
+      "epoch": 0.9531186880658581,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0004244327917864205,
+      "loss": 2.7283,
+      "step": 18691
+    },
+    {
+      "epoch": 0.9531696815219635,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0004239659629640512,
+      "loss": 2.7233,
+      "step": 18692
+    },
+    {
+      "epoch": 0.9532206749780688,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0004234995511534423,
+      "loss": 2.7307,
+      "step": 18693
+    },
+    {
+      "epoch": 0.9532716684341742,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0004230335566422999,
+      "loss": 2.7167,
+      "step": 18694
+    },
+    {
+      "epoch": 0.9533226618902796,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0004225679797180727,
+      "loss": 2.721,
+      "step": 18695
+    },
+    {
+      "epoch": 0.953373655346385,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00042210282066795226,
+      "loss": 2.7172,
+      "step": 18696
+    },
+    {
+      "epoch": 0.9534246488024903,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004216380797788715,
+      "loss": 2.7204,
+      "step": 18697
+    },
+    {
+      "epoch": 0.9534756422585958,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0004211737573375065,
+      "loss": 2.7093,
+      "step": 18698
+    },
+    {
+      "epoch": 0.9535266357147012,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0004207098536302747,
+      "loss": 2.6747,
+      "step": 18699
+    },
+    {
+      "epoch": 0.9535776291708066,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.00042024636894333486,
+      "loss": 2.7074,
+      "step": 18700
+    },
+    {
+      "epoch": 0.953628622626912,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0004197833035625876,
+      "loss": 2.7027,
+      "step": 18701
+    },
+    {
+      "epoch": 0.9536796160830173,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00041932065777367547,
+      "loss": 2.7176,
+      "step": 18702
+    },
+    {
+      "epoch": 0.9537306095391227,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0004188584318619809,
+      "loss": 2.7121,
+      "step": 18703
+    },
+    {
+      "epoch": 0.9537816029952281,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00041839662611262856,
+      "loss": 2.72,
+      "step": 18704
+    },
+    {
+      "epoch": 0.9538325964513336,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00041793524081048297,
+      "loss": 2.7322,
+      "step": 18705
+    },
+    {
+      "epoch": 0.9538835899074389,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0004174742762401503,
+      "loss": 2.6967,
+      "step": 18706
+    },
+    {
+      "epoch": 0.9539345833635443,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00041701373268597644,
+      "loss": 2.7298,
+      "step": 18707
+    },
+    {
+      "epoch": 0.9539855768196497,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00041655361043204785,
+      "loss": 2.7303,
+      "step": 18708
+    },
+    {
+      "epoch": 0.9540365702757551,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0004160939097621909,
+      "loss": 2.7191,
+      "step": 18709
+    },
+    {
+      "epoch": 0.9540875637318604,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0004156346309599722,
+      "loss": 2.6998,
+      "step": 18710
+    },
+    {
+      "epoch": 0.9541385571879658,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00041517577430869794,
+      "loss": 2.7015,
+      "step": 18711
+    },
+    {
+      "epoch": 0.9541895506440713,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00041471734009141395,
+      "loss": 2.7064,
+      "step": 18712
+    },
+    {
+      "epoch": 0.9542405441001767,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0004142593285909053,
+      "loss": 2.7384,
+      "step": 18713
+    },
+    {
+      "epoch": 0.954291537556282,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0004138017400896968,
+      "loss": 2.7154,
+      "step": 18714
+    },
+    {
+      "epoch": 0.9543425310123874,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00041334457487005207,
+      "loss": 2.7193,
+      "step": 18715
+    },
+    {
+      "epoch": 0.9543935244684928,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0004128878332139736,
+      "loss": 2.7386,
+      "step": 18716
+    },
+    {
+      "epoch": 0.9544445179245982,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0004124315154032019,
+      "loss": 2.7201,
+      "step": 18717
+    },
+    {
+      "epoch": 0.9544955113807035,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00041197562171921764,
+      "loss": 2.7142,
+      "step": 18718
+    },
+    {
+      "epoch": 0.954546504836809,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0004115201524432385,
+      "loss": 2.7236,
+      "step": 18719
+    },
+    {
+      "epoch": 0.9545974982929144,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0004110651078562208,
+      "loss": 2.7121,
+      "step": 18720
+    },
+    {
+      "epoch": 0.9546484917490198,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00041061048823885874,
+      "loss": 2.7222,
+      "step": 18721
+    },
+    {
+      "epoch": 0.9546994852051252,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004101562938715853,
+      "loss": 2.7399,
+      "step": 18722
+    },
+    {
+      "epoch": 0.9547504786612305,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0004097025250345696,
+      "loss": 2.7126,
+      "step": 18723
+    },
+    {
+      "epoch": 0.9548014721173359,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0004092491820077193,
+      "loss": 2.7059,
+      "step": 18724
+    },
+    {
+      "epoch": 0.9548524655734413,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.000408796265070679,
+      "loss": 2.7013,
+      "step": 18725
+    },
+    {
+      "epoch": 0.9549034590295468,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004083437745028311,
+      "loss": 2.7076,
+      "step": 18726
+    },
+    {
+      "epoch": 0.9549544524856521,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00040789171058329415,
+      "loss": 2.7348,
+      "step": 18727
+    },
+    {
+      "epoch": 0.9550054459417575,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00040744007359092395,
+      "loss": 2.6983,
+      "step": 18728
+    },
+    {
+      "epoch": 0.9550564393978629,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00040698886380431286,
+      "loss": 2.7274,
+      "step": 18729
+    },
+    {
+      "epoch": 0.9551074328539683,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0004065380815017898,
+      "loss": 2.6863,
+      "step": 18730
+    },
+    {
+      "epoch": 0.9551584263100736,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00040608772696142,
+      "loss": 2.7231,
+      "step": 18731
+    },
+    {
+      "epoch": 0.955209419766179,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00040563780046100464,
+      "loss": 2.7298,
+      "step": 18732
+    },
+    {
+      "epoch": 0.9552604132222845,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0004051883022780808,
+      "loss": 2.7043,
+      "step": 18733
+    },
+    {
+      "epoch": 0.9553114066783899,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.000404739232689922,
+      "loss": 2.7238,
+      "step": 18734
+    },
+    {
+      "epoch": 0.9553624001344952,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0004042905919735367,
+      "loss": 2.7088,
+      "step": 18735
+    },
+    {
+      "epoch": 0.9554133935906006,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00040384238040566916,
+      "loss": 2.7193,
+      "step": 18736
+    },
+    {
+      "epoch": 0.955464387046706,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0004033945982627986,
+      "loss": 2.7561,
+      "step": 18737
+    },
+    {
+      "epoch": 0.9555153805028114,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00040294724582113966,
+      "loss": 2.741,
+      "step": 18738
+    },
+    {
+      "epoch": 0.9555663739589167,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0004025003233566418,
+      "loss": 2.7418,
+      "step": 18739
+    },
+    {
+      "epoch": 0.9556173674150222,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00040205383114498913,
+      "loss": 2.7224,
+      "step": 18740
+    },
+    {
+      "epoch": 0.9556683608711276,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0004016077694616006,
+      "loss": 2.7404,
+      "step": 18741
+    },
+    {
+      "epoch": 0.955719354327233,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0004011621385816292,
+      "loss": 2.7195,
+      "step": 18742
+    },
+    {
+      "epoch": 0.9557703477833384,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00040071693877996296,
+      "loss": 2.7328,
+      "step": 18743
+    },
+    {
+      "epoch": 0.9558213412394437,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004002721703312232,
+      "loss": 2.7043,
+      "step": 18744
+    },
+    {
+      "epoch": 0.9558723346955491,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0003998278335097657,
+      "loss": 2.725,
+      "step": 18745
+    },
+    {
+      "epoch": 0.9559233281516545,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00039938392858967895,
+      "loss": 2.7313,
+      "step": 18746
+    },
+    {
+      "epoch": 0.95597432160776,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00039894045584478664,
+      "loss": 2.7003,
+      "step": 18747
+    },
+    {
+      "epoch": 0.9560253150638653,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003984974155486447,
+      "loss": 2.6927,
+      "step": 18748
+    },
+    {
+      "epoch": 0.9560763085199707,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00039805480797454286,
+      "loss": 2.7393,
+      "step": 18749
+    },
+    {
+      "epoch": 0.9561273019760761,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00039761263339550324,
+      "loss": 2.6838,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9561782954321815,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003971708920842818,
+      "loss": 2.6968,
+      "step": 18751
+    },
+    {
+      "epoch": 0.9562292888882868,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00039672958431336686,
+      "loss": 2.7392,
+      "step": 18752
+    },
+    {
+      "epoch": 0.9562802823443922,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00039628871035497884,
+      "loss": 2.7096,
+      "step": 18753
+    },
+    {
+      "epoch": 0.9563312758004977,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003958482704810707,
+      "loss": 2.7049,
+      "step": 18754
+    },
+    {
+      "epoch": 0.9563822692566031,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003954082649633287,
+      "loss": 2.7253,
+      "step": 18755
+    },
+    {
+      "epoch": 0.9564332627127085,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00039496869407316975,
+      "loss": 2.7122,
+      "step": 18756
+    },
+    {
+      "epoch": 0.9564842561688138,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00039452955808174354,
+      "loss": 2.6982,
+      "step": 18757
+    },
+    {
+      "epoch": 0.9565352496249192,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003940908572599308,
+      "loss": 2.7291,
+      "step": 18758
+    },
+    {
+      "epoch": 0.9565862430810246,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00039365259187834514,
+      "loss": 2.7469,
+      "step": 18759
+    },
+    {
+      "epoch": 0.95663723653713,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00039321476220732977,
+      "loss": 2.7042,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9566882299932354,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00039277736851696035,
+      "loss": 2.7208,
+      "step": 18761
+    },
+    {
+      "epoch": 0.9567392234493408,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000392340411077043,
+      "loss": 2.695,
+      "step": 18762
+    },
+    {
+      "epoch": 0.9567902169054462,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003919038901571156,
+      "loss": 2.703,
+      "step": 18763
+    },
+    {
+      "epoch": 0.9568412103615516,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003914678060264457,
+      "loss": 2.7438,
+      "step": 18764
+    },
+    {
+      "epoch": 0.9568922038176569,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003910321589540322,
+      "loss": 2.7194,
+      "step": 18765
+    },
+    {
+      "epoch": 0.9569431972737623,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00039059694920860413,
+      "loss": 2.6989,
+      "step": 18766
+    },
+    {
+      "epoch": 0.9569941907298677,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0003901621770586203,
+      "loss": 2.7066,
+      "step": 18767
+    },
+    {
+      "epoch": 0.9570451841859732,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0003897278427722705,
+      "loss": 2.7362,
+      "step": 18768
+    },
+    {
+      "epoch": 0.9570961776420785,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003892939466174737,
+      "loss": 2.726,
+      "step": 18769
+    },
+    {
+      "epoch": 0.9571471710981839,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00038886048886187865,
+      "loss": 2.7187,
+      "step": 18770
+    },
+    {
+      "epoch": 0.9571981645542893,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003884274697728644,
+      "loss": 2.7319,
+      "step": 18771
+    },
+    {
+      "epoch": 0.9572491580103947,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00038799488961753845,
+      "loss": 2.7281,
+      "step": 18772
+    },
+    {
+      "epoch": 0.9573001514665,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0003875627486627382,
+      "loss": 2.7245,
+      "step": 18773
+    },
+    {
+      "epoch": 0.9573511449226054,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0003871310471750297,
+      "loss": 2.7004,
+      "step": 18774
+    },
+    {
+      "epoch": 0.9574021383787109,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003866997854207084,
+      "loss": 2.7145,
+      "step": 18775
+    },
+    {
+      "epoch": 0.9574531318348163,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0003862689636657979,
+      "loss": 2.7154,
+      "step": 18776
+    },
+    {
+      "epoch": 0.9575041252909217,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0003858385821760511,
+      "loss": 2.702,
+      "step": 18777
+    },
+    {
+      "epoch": 0.957555118747027,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003854086412169482,
+      "loss": 2.7024,
+      "step": 18778
+    },
+    {
+      "epoch": 0.9576061122031324,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00038497914105369913,
+      "loss": 2.6917,
+      "step": 18779
+    },
+    {
+      "epoch": 0.9576571056592378,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.000384550081951241,
+      "loss": 2.7258,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9577080991153432,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00038412146417423905,
+      "loss": 2.7064,
+      "step": 18781
+    },
+    {
+      "epoch": 0.9577590925714486,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0003836932879870858,
+      "loss": 2.7212,
+      "step": 18782
+    },
+    {
+      "epoch": 0.957810086027554,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00038326555365390233,
+      "loss": 2.706,
+      "step": 18783
+    },
+    {
+      "epoch": 0.9578610794836594,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0003828382614385365,
+      "loss": 2.7284,
+      "step": 18784
+    },
+    {
+      "epoch": 0.9579120729397648,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00038241141160456373,
+      "loss": 2.7266,
+      "step": 18785
+    },
+    {
+      "epoch": 0.9579630663958701,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00038198500441528606,
+      "loss": 2.684,
+      "step": 18786
+    },
+    {
+      "epoch": 0.9580140598519755,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0003815590401337333,
+      "loss": 2.7252,
+      "step": 18787
+    },
+    {
+      "epoch": 0.9580650533080809,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00038113351902266184,
+      "loss": 2.6948,
+      "step": 18788
+    },
+    {
+      "epoch": 0.9581160467641864,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00038070844134455384,
+      "loss": 2.7333,
+      "step": 18789
+    },
+    {
+      "epoch": 0.9581670402202918,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003802838073616185,
+      "loss": 2.7196,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9582180336763971,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003798596173357921,
+      "loss": 2.7109,
+      "step": 18791
+    },
+    {
+      "epoch": 0.9582690271325025,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0003794358715287359,
+      "loss": 2.6858,
+      "step": 18792
+    },
+    {
+      "epoch": 0.9583200205886079,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00037901257020183786,
+      "loss": 2.7334,
+      "step": 18793
+    },
+    {
+      "epoch": 0.9583710140447133,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00037858971361621097,
+      "loss": 2.7127,
+      "step": 18794
+    },
+    {
+      "epoch": 0.9584220075008186,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00037816730203269535,
+      "loss": 2.7257,
+      "step": 18795
+    },
+    {
+      "epoch": 0.9584730009569241,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003777453357118548,
+      "loss": 2.7333,
+      "step": 18796
+    },
+    {
+      "epoch": 0.9585239944130295,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003773238149139798,
+      "loss": 2.7399,
+      "step": 18797
+    },
+    {
+      "epoch": 0.9585749878691349,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003769027398990852,
+      "loss": 2.6962,
+      "step": 18798
+    },
+    {
+      "epoch": 0.9586259813252402,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003764821109269115,
+      "loss": 2.7321,
+      "step": 18799
+    },
+    {
+      "epoch": 0.9586769747813456,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000376061928256924,
+      "loss": 2.7066,
+      "step": 18800
+    },
+    {
+      "epoch": 0.958727968237451,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0003756421921483122,
+      "loss": 2.6817,
+      "step": 18801
+    },
+    {
+      "epoch": 0.9587789616935564,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003752229028599903,
+      "loss": 2.7114,
+      "step": 18802
+    },
+    {
+      "epoch": 0.9588299551496619,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037480406065059724,
+      "loss": 2.7103,
+      "step": 18803
+    },
+    {
+      "epoch": 0.9588809486057672,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00037438566577849584,
+      "loss": 2.7204,
+      "step": 18804
+    },
+    {
+      "epoch": 0.9589319420618726,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0003739677185017732,
+      "loss": 2.7409,
+      "step": 18805
+    },
+    {
+      "epoch": 0.958982935517978,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003735502190782397,
+      "loss": 2.7057,
+      "step": 18806
+    },
+    {
+      "epoch": 0.9590339289740833,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003731331677654307,
+      "loss": 2.7136,
+      "step": 18807
+    },
+    {
+      "epoch": 0.9590849224301887,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003727165648206039,
+      "loss": 2.7409,
+      "step": 18808
+    },
+    {
+      "epoch": 0.9591359158862941,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003723004105007412,
+      "loss": 2.7207,
+      "step": 18809
+    },
+    {
+      "epoch": 0.9591869093423996,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00037188470506254747,
+      "loss": 2.7264,
+      "step": 18810
+    },
+    {
+      "epoch": 0.959237902798505,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00037146944876245046,
+      "loss": 2.6907,
+      "step": 18811
+    },
+    {
+      "epoch": 0.9592888962546103,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0003710546418566013,
+      "loss": 2.7379,
+      "step": 18812
+    },
+    {
+      "epoch": 0.9593398897107157,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00037064028460087367,
+      "loss": 2.7439,
+      "step": 18813
+    },
+    {
+      "epoch": 0.9593908831668211,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0003702263772508638,
+      "loss": 2.7326,
+      "step": 18814
+    },
+    {
+      "epoch": 0.9594418766229265,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00036981292006189095,
+      "loss": 2.728,
+      "step": 18815
+    },
+    {
+      "epoch": 0.9594928700790318,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00036939991328899615,
+      "loss": 2.7023,
+      "step": 18816
+    },
+    {
+      "epoch": 0.9595438635351373,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00036898735718694255,
+      "loss": 2.7244,
+      "step": 18817
+    },
+    {
+      "epoch": 0.9595948569912427,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0003685752520102155,
+      "loss": 2.7369,
+      "step": 18818
+    },
+    {
+      "epoch": 0.9596458504473481,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00036816359801302226,
+      "loss": 2.7246,
+      "step": 18819
+    },
+    {
+      "epoch": 0.9596968439034534,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003677523954492918,
+      "loss": 2.6949,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9597478373595588,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00036734164457267435,
+      "loss": 2.7141,
+      "step": 18821
+    },
+    {
+      "epoch": 0.9597988308156642,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0003669313456365415,
+      "loss": 2.7274,
+      "step": 18822
+    },
+    {
+      "epoch": 0.9598498242717696,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00036652149889398675,
+      "loss": 2.7251,
+      "step": 18823
+    },
+    {
+      "epoch": 0.959900817727875,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003661121045978241,
+      "loss": 2.7287,
+      "step": 18824
+    },
+    {
+      "epoch": 0.9599518111839804,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00036570316300058806,
+      "loss": 2.7011,
+      "step": 18825
+    },
+    {
+      "epoch": 0.9600028046400858,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00036529467435453446,
+      "loss": 2.7377,
+      "step": 18826
+    },
+    {
+      "epoch": 0.9600537980961912,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00036488663891164,
+      "loss": 2.7269,
+      "step": 18827
+    },
+    {
+      "epoch": 0.9601047915522966,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00036447905692360127,
+      "loss": 2.7092,
+      "step": 18828
+    },
+    {
+      "epoch": 0.9601557850084019,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00036407192864183513,
+      "loss": 2.702,
+      "step": 18829
+    },
+    {
+      "epoch": 0.9602067784645073,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00036366525431747875,
+      "loss": 2.7172,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9602577719206128,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00036325903420138993,
+      "loss": 2.7158,
+      "step": 18831
+    },
+    {
+      "epoch": 0.9603087653767182,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00036285326854414504,
+      "loss": 2.6955,
+      "step": 18832
+    },
+    {
+      "epoch": 0.9603597588328235,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036244795759604056,
+      "loss": 2.7372,
+      "step": 18833
+    },
+    {
+      "epoch": 0.9604107522889289,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00036204310160709285,
+      "loss": 2.7329,
+      "step": 18834
+    },
+    {
+      "epoch": 0.9604617457450343,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00036163870082703774,
+      "loss": 2.7007,
+      "step": 18835
+    },
+    {
+      "epoch": 0.9605127392011397,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00036123475550532973,
+      "loss": 2.7135,
+      "step": 18836
+    },
+    {
+      "epoch": 0.960563732657245,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00036083126589114253,
+      "loss": 2.7078,
+      "step": 18837
+    },
+    {
+      "epoch": 0.9606147261133505,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003604282322333689,
+      "loss": 2.73,
+      "step": 18838
+    },
+    {
+      "epoch": 0.9606657195694559,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0003600256547806201,
+      "loss": 2.7519,
+      "step": 18839
+    },
+    {
+      "epoch": 0.9607167130255613,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0003596235337812263,
+      "loss": 2.7199,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9607677064816667,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00035922186948323585,
+      "loss": 2.7141,
+      "step": 18841
+    },
+    {
+      "epoch": 0.960818699937772,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003588206621344153,
+      "loss": 2.7449,
+      "step": 18842
+    },
+    {
+      "epoch": 0.9608696933938774,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00035841991198224977,
+      "loss": 2.6933,
+      "step": 18843
+    },
+    {
+      "epoch": 0.9609206868499828,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00035801961927394224,
+      "loss": 2.7193,
+      "step": 18844
+    },
+    {
+      "epoch": 0.9609716803060883,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00035761978425641295,
+      "loss": 2.7515,
+      "step": 18845
+    },
+    {
+      "epoch": 0.9610226737621936,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00035722040717630043,
+      "loss": 2.7222,
+      "step": 18846
+    },
+    {
+      "epoch": 0.961073667218299,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0003568214882799606,
+      "loss": 2.7224,
+      "step": 18847
+    },
+    {
+      "epoch": 0.9611246606744044,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00035642302781346645,
+      "loss": 2.7497,
+      "step": 18848
+    },
+    {
+      "epoch": 0.9611756541305098,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00035602502602260854,
+      "loss": 2.7092,
+      "step": 18849
+    },
+    {
+      "epoch": 0.9612266475866151,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003556274831528943,
+      "loss": 2.7446,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9612776410427205,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00035523039944954836,
+      "loss": 2.7105,
+      "step": 18851
+    },
+    {
+      "epoch": 0.961328634498826,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00035483377515751173,
+      "loss": 2.7009,
+      "step": 18852
+    },
+    {
+      "epoch": 0.9613796279549314,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0003544376105214424,
+      "loss": 2.7022,
+      "step": 18853
+    },
+    {
+      "epoch": 0.9614306214110367,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0003540419057857143,
+      "loss": 2.7182,
+      "step": 18854
+    },
+    {
+      "epoch": 0.9614816148671421,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003536466611944182,
+      "loss": 2.7257,
+      "step": 18855
+    },
+    {
+      "epoch": 0.9615326083232475,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00035325187699136084,
+      "loss": 2.7289,
+      "step": 18856
+    },
+    {
+      "epoch": 0.9615836017793529,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00035285755342006475,
+      "loss": 2.7001,
+      "step": 18857
+    },
+    {
+      "epoch": 0.9616345952354582,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003524636907237685,
+      "loss": 2.7247,
+      "step": 18858
+    },
+    {
+      "epoch": 0.9616855886915637,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00035207028914542685,
+      "loss": 2.7194,
+      "step": 18859
+    },
+    {
+      "epoch": 0.9617365821476691,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003516773489277092,
+      "loss": 2.7129,
+      "step": 18860
+    },
+    {
+      "epoch": 0.9617875756037745,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003512848703130014,
+      "loss": 2.6893,
+      "step": 18861
+    },
+    {
+      "epoch": 0.9618385690598799,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00035089285354340306,
+      "loss": 2.6776,
+      "step": 18862
+    },
+    {
+      "epoch": 0.9618895625159852,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00035050129886073035,
+      "loss": 2.7211,
+      "step": 18863
+    },
+    {
+      "epoch": 0.9619405559720906,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0003501102065065139,
+      "loss": 2.6983,
+      "step": 18864
+    },
+    {
+      "epoch": 0.961991549428196,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003497195767219991,
+      "loss": 2.7505,
+      "step": 18865
+    },
+    {
+      "epoch": 0.9620425428843015,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00034932940974814595,
+      "loss": 2.7167,
+      "step": 18866
+    },
+    {
+      "epoch": 0.9620935363404068,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00034893970582562916,
+      "loss": 2.6845,
+      "step": 18867
+    },
+    {
+      "epoch": 0.9621445297965122,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0003485504651948376,
+      "loss": 2.7174,
+      "step": 18868
+    },
+    {
+      "epoch": 0.9621955232526176,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0003481616880958745,
+      "loss": 2.7096,
+      "step": 18869
+    },
+    {
+      "epoch": 0.962246516708723,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00034777337476855696,
+      "loss": 2.7315,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9622975101648283,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003473855254524161,
+      "loss": 2.7132,
+      "step": 18871
+    },
+    {
+      "epoch": 0.9623485036209337,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00034699814038669717,
+      "loss": 2.6974,
+      "step": 18872
+    },
+    {
+      "epoch": 0.9623994970770392,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003466112198103587,
+      "loss": 2.6846,
+      "step": 18873
+    },
+    {
+      "epoch": 0.9624504905331446,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00034622476396207254,
+      "loss": 2.7263,
+      "step": 18874
+    },
+    {
+      "epoch": 0.96250148398925,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00034583877308022424,
+      "loss": 2.7216,
+      "step": 18875
+    },
+    {
+      "epoch": 0.9625524774453553,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00034545324740291224,
+      "loss": 2.7099,
+      "step": 18876
+    },
+    {
+      "epoch": 0.9626034709014607,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003450681871679482,
+      "loss": 2.718,
+      "step": 18877
+    },
+    {
+      "epoch": 0.9626544643575661,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00034468359261285666,
+      "loss": 2.7021,
+      "step": 18878
+    },
+    {
+      "epoch": 0.9627054578136715,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00034429946397487464,
+      "loss": 2.7028,
+      "step": 18879
+    },
+    {
+      "epoch": 0.9627564512697769,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003439158014909523,
+      "loss": 2.711,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9628074447258823,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003435326053977519,
+      "loss": 2.7363,
+      "step": 18881
+    },
+    {
+      "epoch": 0.9628584381819877,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0003431498759316479,
+      "loss": 2.7074,
+      "step": 18882
+    },
+    {
+      "epoch": 0.9629094316380931,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00034276761332872706,
+      "loss": 2.7152,
+      "step": 18883
+    },
+    {
+      "epoch": 0.9629604250941984,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00034238581782478837,
+      "loss": 2.7325,
+      "step": 18884
+    },
+    {
+      "epoch": 0.9630114185503038,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00034200448965534223,
+      "loss": 2.7233,
+      "step": 18885
+    },
+    {
+      "epoch": 0.9630624120064092,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0003416236290556112,
+      "loss": 2.6994,
+      "step": 18886
+    },
+    {
+      "epoch": 0.9631134054625147,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0003412432362605288,
+      "loss": 2.7025,
+      "step": 18887
+    },
+    {
+      "epoch": 0.96316439891862,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000340863311504741,
+      "loss": 2.7284,
+      "step": 18888
+    },
+    {
+      "epoch": 0.9632153923747254,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0003404838550226043,
+      "loss": 2.7263,
+      "step": 18889
+    },
+    {
+      "epoch": 0.9632663858308308,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00034010486704818665,
+      "loss": 2.7134,
+      "step": 18890
+    },
+    {
+      "epoch": 0.9633173792869362,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00033972634781526633,
+      "loss": 2.7101,
+      "step": 18891
+    },
+    {
+      "epoch": 0.9633683727430415,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00033934829755733354,
+      "loss": 2.7163,
+      "step": 18892
+    },
+    {
+      "epoch": 0.9634193661991469,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003389707165075884,
+      "loss": 2.7072,
+      "step": 18893
+    },
+    {
+      "epoch": 0.9634703596552524,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003385936048989421,
+      "loss": 2.7259,
+      "step": 18894
+    },
+    {
+      "epoch": 0.9635213531113578,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003382169629640157,
+      "loss": 2.7166,
+      "step": 18895
+    },
+    {
+      "epoch": 0.9635723465674632,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00033784079093514133,
+      "loss": 2.7353,
+      "step": 18896
+    },
+    {
+      "epoch": 0.9636233400235685,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003374650890443607,
+      "loss": 2.7199,
+      "step": 18897
+    },
+    {
+      "epoch": 0.9636743334796739,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003370898575234253,
+      "loss": 2.7227,
+      "step": 18898
+    },
+    {
+      "epoch": 0.9637253269357793,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00033671509660379676,
+      "loss": 2.72,
+      "step": 18899
+    },
+    {
+      "epoch": 0.9637763203918847,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0003363408065166468,
+      "loss": 2.7263,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9638273138479901,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00033596698749285614,
+      "loss": 2.7116,
+      "step": 18901
+    },
+    {
+      "epoch": 0.9638783073040955,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00033559363976301516,
+      "loss": 2.713,
+      "step": 18902
+    },
+    {
+      "epoch": 0.9639293007602009,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00033522076355742357,
+      "loss": 2.6863,
+      "step": 18903
+    },
+    {
+      "epoch": 0.9639802942163063,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00033484835910609014,
+      "loss": 2.7398,
+      "step": 18904
+    },
+    {
+      "epoch": 0.9640312876724116,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0003344764266387326,
+      "loss": 2.7119,
+      "step": 18905
+    },
+    {
+      "epoch": 0.964082281128517,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0003341049663847775,
+      "loss": 2.6781,
+      "step": 18906
+    },
+    {
+      "epoch": 0.9641332745846224,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00033373397857336005,
+      "loss": 2.6934,
+      "step": 18907
+    },
+    {
+      "epoch": 0.9641842680407279,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003333634634333247,
+      "loss": 2.7013,
+      "step": 18908
+    },
+    {
+      "epoch": 0.9642352614968333,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0003329934211932235,
+      "loss": 2.7008,
+      "step": 18909
+    },
+    {
+      "epoch": 0.9642862549529386,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00033262385208131706,
+      "loss": 2.7207,
+      "step": 18910
+    },
+    {
+      "epoch": 0.964337248409044,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00033225475632557425,
+      "loss": 2.7355,
+      "step": 18911
+    },
+    {
+      "epoch": 0.9643882418651494,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00033188613415367183,
+      "loss": 2.7612,
+      "step": 18912
+    },
+    {
+      "epoch": 0.9644392353212548,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0003315179857929944,
+      "loss": 2.7238,
+      "step": 18913
+    },
+    {
+      "epoch": 0.9644902287773601,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00033115031147063444,
+      "loss": 2.7154,
+      "step": 18914
+    },
+    {
+      "epoch": 0.9645412222334656,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0003307831114133919,
+      "loss": 2.7094,
+      "step": 18915
+    },
+    {
+      "epoch": 0.964592215689571,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00033041638584777446,
+      "loss": 2.725,
+      "step": 18916
+    },
+    {
+      "epoch": 0.9646432091456764,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0003300501349999966,
+      "loss": 2.7137,
+      "step": 18917
+    },
+    {
+      "epoch": 0.9646942026017817,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0003296843590959804,
+      "loss": 2.7068,
+      "step": 18918
+    },
+    {
+      "epoch": 0.9647451960578871,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00032931905836135466,
+      "loss": 2.7251,
+      "step": 18919
+    },
+    {
+      "epoch": 0.9647961895139925,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00032895423302145544,
+      "loss": 2.676,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9648471829700979,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003285898833013252,
+      "loss": 2.707,
+      "step": 18921
+    },
+    {
+      "epoch": 0.9648981764262033,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003282260094257131,
+      "loss": 2.6953,
+      "step": 18922
+    },
+    {
+      "epoch": 0.9649491698823087,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003278626116190747,
+      "loss": 2.7245,
+      "step": 18923
+    },
+    {
+      "epoch": 0.9650001633384141,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00032749969010557254,
+      "loss": 2.7235,
+      "step": 18924
+    },
+    {
+      "epoch": 0.9650511567945195,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003271372451090745,
+      "loss": 2.736,
+      "step": 18925
+    },
+    {
+      "epoch": 0.9651021502506248,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0003267752768531551,
+      "loss": 2.7007,
+      "step": 18926
+    },
+    {
+      "epoch": 0.9651531437067302,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00032641378556109394,
+      "loss": 2.7268,
+      "step": 18927
+    },
+    {
+      "epoch": 0.9652041371628356,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00032605277145587755,
+      "loss": 2.7408,
+      "step": 18928
+    },
+    {
+      "epoch": 0.9652551306189411,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0003256922347601975,
+      "loss": 2.705,
+      "step": 18929
+    },
+    {
+      "epoch": 0.9653061240750465,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00032533217569645083,
+      "loss": 2.7077,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9653571175311518,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0003249725944867399,
+      "loss": 2.7066,
+      "step": 18931
+    },
+    {
+      "epoch": 0.9654081109872572,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00032461349135287296,
+      "loss": 2.72,
+      "step": 18932
+    },
+    {
+      "epoch": 0.9654591044433626,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0003242548665163627,
+      "loss": 2.705,
+      "step": 18933
+    },
+    {
+      "epoch": 0.965510097899468,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003238967201984266,
+      "loss": 2.7009,
+      "step": 18934
+    },
+    {
+      "epoch": 0.9655610913555733,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003235390526199874,
+      "loss": 2.7143,
+      "step": 18935
+    },
+    {
+      "epoch": 0.9656120848116788,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00032318186400167264,
+      "loss": 2.7074,
+      "step": 18936
+    },
+    {
+      "epoch": 0.9656630782677842,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003228251545638143,
+      "loss": 2.6915,
+      "step": 18937
+    },
+    {
+      "epoch": 0.9657140717238896,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00032246892452644825,
+      "loss": 2.7026,
+      "step": 18938
+    },
+    {
+      "epoch": 0.9657650651799949,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003221131741093153,
+      "loss": 2.722,
+      "step": 18939
+    },
+    {
+      "epoch": 0.9658160586361003,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00032175790353186054,
+      "loss": 2.7057,
+      "step": 18940
+    },
+    {
+      "epoch": 0.9658670520922057,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0003214031130132323,
+      "loss": 2.7328,
+      "step": 18941
+    },
+    {
+      "epoch": 0.9659180455483111,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003210488027722831,
+      "loss": 2.6952,
+      "step": 18942
+    },
+    {
+      "epoch": 0.9659690390044166,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00032069497302756933,
+      "loss": 2.7376,
+      "step": 18943
+    },
+    {
+      "epoch": 0.9660200324605219,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0003203416239973513,
+      "loss": 2.7323,
+      "step": 18944
+    },
+    {
+      "epoch": 0.9660710259166273,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003199887558995923,
+      "loss": 2.7429,
+      "step": 18945
+    },
+    {
+      "epoch": 0.9661220193727327,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00031963636895195903,
+      "loss": 2.6988,
+      "step": 18946
+    },
+    {
+      "epoch": 0.966173012828838,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0003192844633718214,
+      "loss": 2.6928,
+      "step": 18947
+    },
+    {
+      "epoch": 0.9662240062849434,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00031893303937625267,
+      "loss": 2.7385,
+      "step": 18948
+    },
+    {
+      "epoch": 0.9662749997410488,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003185820971820285,
+      "loss": 2.7221,
+      "step": 18949
+    },
+    {
+      "epoch": 0.9663259931971543,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0003182316370056281,
+      "loss": 2.7108,
+      "step": 18950
+    },
+    {
+      "epoch": 0.9663769866532597,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003178816590632323,
+      "loss": 2.7079,
+      "step": 18951
+    },
+    {
+      "epoch": 0.966427980109365,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003175321635707256,
+      "loss": 2.7127,
+      "step": 18952
+    },
+    {
+      "epoch": 0.9664789735654704,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00031718315074369424,
+      "loss": 2.711,
+      "step": 18953
+    },
+    {
+      "epoch": 0.9665299670215758,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0003168346207974269,
+      "loss": 2.7245,
+      "step": 18954
+    },
+    {
+      "epoch": 0.9665809604776812,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003164865739469144,
+      "loss": 2.71,
+      "step": 18955
+    },
+    {
+      "epoch": 0.9666319539337865,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00031613901040684927,
+      "loss": 2.7076,
+      "step": 18956
+    },
+    {
+      "epoch": 0.966682947389892,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003157919303916265,
+      "loss": 2.6969,
+      "step": 18957
+    },
+    {
+      "epoch": 0.9667339408459974,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00031544533411534214,
+      "loss": 2.7012,
+      "step": 18958
+    },
+    {
+      "epoch": 0.9667849343021028,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0003150992217917942,
+      "loss": 2.6967,
+      "step": 18959
+    },
+    {
+      "epoch": 0.9668359277582081,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0003147535936344826,
+      "loss": 2.7267,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9668869212143135,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00031440844985660776,
+      "loss": 2.7009,
+      "step": 18961
+    },
+    {
+      "epoch": 0.9669379146704189,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003140637906710721,
+      "loss": 2.7168,
+      "step": 18962
+    },
+    {
+      "epoch": 0.9669889081265243,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00031371961629047797,
+      "loss": 2.7454,
+      "step": 18963
+    },
+    {
+      "epoch": 0.9670399015826298,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0003133759269271301,
+      "loss": 2.6861,
+      "step": 18964
+    },
+    {
+      "epoch": 0.9670908950387351,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00031303272279303314,
+      "loss": 2.6748,
+      "step": 18965
+    },
+    {
+      "epoch": 0.9671418884948405,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0003126900040998926,
+      "loss": 2.708,
+      "step": 18966
+    },
+    {
+      "epoch": 0.9671928819509459,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0003123477710591145,
+      "loss": 2.7171,
+      "step": 18967
+    },
+    {
+      "epoch": 0.9672438754070513,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003120060238818056,
+      "loss": 2.7102,
+      "step": 18968
+    },
+    {
+      "epoch": 0.9672948688631566,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003116647627787729,
+      "loss": 2.72,
+      "step": 18969
+    },
+    {
+      "epoch": 0.967345862319262,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000311323987960523,
+      "loss": 2.7036,
+      "step": 18970
+    },
+    {
+      "epoch": 0.9673968557753675,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003109836996372627,
+      "loss": 2.7347,
+      "step": 18971
+    },
+    {
+      "epoch": 0.9674478492314729,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00031064389801889957,
+      "loss": 2.6885,
+      "step": 18972
+    },
+    {
+      "epoch": 0.9674988426875782,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00031030458331504006,
+      "loss": 2.685,
+      "step": 18973
+    },
+    {
+      "epoch": 0.9675498361436836,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0003099657557349906,
+      "loss": 2.6914,
+      "step": 18974
+    },
+    {
+      "epoch": 0.967600829599789,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003096274154877567,
+      "loss": 2.6928,
+      "step": 18975
+    },
+    {
+      "epoch": 0.9676518230558944,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00030928956278204425,
+      "loss": 2.6845,
+      "step": 18976
+    },
+    {
+      "epoch": 0.9677028165119997,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0003089521978262572,
+      "loss": 2.7081,
+      "step": 18977
+    },
+    {
+      "epoch": 0.9677538099681052,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00030861532082849955,
+      "loss": 2.703,
+      "step": 18978
+    },
+    {
+      "epoch": 0.9678048034242106,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00030827893199657355,
+      "loss": 2.7294,
+      "step": 18979
+    },
+    {
+      "epoch": 0.967855796880316,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00030794303153798124,
+      "loss": 2.7198,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9679067903364214,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003076076196599227,
+      "loss": 2.7109,
+      "step": 18981
+    },
+    {
+      "epoch": 0.9679577837925267,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003072726965692968,
+      "loss": 2.7241,
+      "step": 18982
+    },
+    {
+      "epoch": 0.9680087772486321,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00030693826247270094,
+      "loss": 2.7019,
+      "step": 18983
+    },
+    {
+      "epoch": 0.9680597707047375,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0003066043175764309,
+      "loss": 2.7246,
+      "step": 18984
+    },
+    {
+      "epoch": 0.968110764160843,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0003062708620864806,
+      "loss": 2.6918,
+      "step": 18985
+    },
+    {
+      "epoch": 0.9681617576169483,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00030593789620854216,
+      "loss": 2.7255,
+      "step": 18986
+    },
+    {
+      "epoch": 0.9682127510730537,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00030560542014800566,
+      "loss": 2.7099,
+      "step": 18987
+    },
+    {
+      "epoch": 0.9682637445291591,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00030527343410995903,
+      "loss": 2.7241,
+      "step": 18988
+    },
+    {
+      "epoch": 0.9683147379852645,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0003049419382991881,
+      "loss": 2.7222,
+      "step": 18989
+    },
+    {
+      "epoch": 0.9683657314413698,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00030461093292017587,
+      "loss": 2.6818,
+      "step": 18990
+    },
+    {
+      "epoch": 0.9684167248974752,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003042804181771035,
+      "loss": 2.691,
+      "step": 18991
+    },
+    {
+      "epoch": 0.9684677183535807,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00030395039427384864,
+      "loss": 2.6947,
+      "step": 18992
+    },
+    {
+      "epoch": 0.9685187118096861,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00030362086141398674,
+      "loss": 2.7148,
+      "step": 18993
+    },
+    {
+      "epoch": 0.9685697052657914,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003032918198007904,
+      "loss": 2.7307,
+      "step": 18994
+    },
+    {
+      "epoch": 0.9686206987218968,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00030296326963722895,
+      "loss": 2.7324,
+      "step": 18995
+    },
+    {
+      "epoch": 0.9686716921780022,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00030263521112596826,
+      "loss": 2.7374,
+      "step": 18996
+    },
+    {
+      "epoch": 0.9687226856341076,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003023076444693719,
+      "loss": 2.6912,
+      "step": 18997
+    },
+    {
+      "epoch": 0.968773679090213,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0003019805698694991,
+      "loss": 2.7216,
+      "step": 18998
+    },
+    {
+      "epoch": 0.9688246725463184,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003016539875281063,
+      "loss": 2.6952,
+      "step": 18999
+    },
+    {
+      "epoch": 0.9688756660024238,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0003013278976466452,
+      "loss": 2.6973,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9689266594585292,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0003010023004262649,
+      "loss": 2.7022,
+      "step": 19001
+    },
+    {
+      "epoch": 0.9689776529146346,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00030067719606780995,
+      "loss": 2.7247,
+      "step": 19002
+    },
+    {
+      "epoch": 0.9690286463707399,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0003003525847718214,
+      "loss": 2.7007,
+      "step": 19003
+    },
+    {
+      "epoch": 0.9690796398268453,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00030002846673853506,
+      "loss": 2.7023,
+      "step": 19004
+    },
+    {
+      "epoch": 0.9691306332829507,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002997048421678842,
+      "loss": 2.7105,
+      "step": 19005
+    },
+    {
+      "epoch": 0.9691816267390562,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.000299381711259496,
+      "loss": 2.7275,
+      "step": 19006
+    },
+    {
+      "epoch": 0.9692326201951615,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00029905907421269405,
+      "loss": 2.6898,
+      "step": 19007
+    },
+    {
+      "epoch": 0.9692836136512669,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002987369312264969,
+      "loss": 2.7332,
+      "step": 19008
+    },
+    {
+      "epoch": 0.9693346071073723,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00029841528249961887,
+      "loss": 2.724,
+      "step": 19009
+    },
+    {
+      "epoch": 0.9693856005634777,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000298094128230469,
+      "loss": 2.7118,
+      "step": 19010
+    },
+    {
+      "epoch": 0.969436594019583,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00029777346861715125,
+      "loss": 2.6998,
+      "step": 19011
+    },
+    {
+      "epoch": 0.9694875874756884,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00029745330385746457,
+      "loss": 2.7193,
+      "step": 19012
+    },
+    {
+      "epoch": 0.9695385809317939,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00029713363414890276,
+      "loss": 2.7231,
+      "step": 19013
+    },
+    {
+      "epoch": 0.9695895743878993,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000296814459688654,
+      "loss": 2.7429,
+      "step": 19014
+    },
+    {
+      "epoch": 0.9696405678440047,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002964957806736013,
+      "loss": 2.7041,
+      "step": 19015
+    },
+    {
+      "epoch": 0.96969156130011,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0002961775973003217,
+      "loss": 2.728,
+      "step": 19016
+    },
+    {
+      "epoch": 0.9697425547562154,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.000295859909765087,
+      "loss": 2.7109,
+      "step": 19017
+    },
+    {
+      "epoch": 0.9697935482123208,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00029554271826386265,
+      "loss": 2.7206,
+      "step": 19018
+    },
+    {
+      "epoch": 0.9698445416684262,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00029522602299230836,
+      "loss": 2.6923,
+      "step": 19019
+    },
+    {
+      "epoch": 0.9698955351245316,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0002949098241457775,
+      "loss": 2.7365,
+      "step": 19020
+    },
+    {
+      "epoch": 0.969946528580637,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00029459412191931765,
+      "loss": 2.7103,
+      "step": 19021
+    },
+    {
+      "epoch": 0.9699975220367424,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0002942789165076696,
+      "loss": 2.739,
+      "step": 19022
+    },
+    {
+      "epoch": 0.9700485154928478,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0002939642081052681,
+      "loss": 2.7146,
+      "step": 19023
+    },
+    {
+      "epoch": 0.9700995089489531,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00029364999690624066,
+      "loss": 2.727,
+      "step": 19024
+    },
+    {
+      "epoch": 0.9701505024050585,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00029333628310440916,
+      "loss": 2.7325,
+      "step": 19025
+    },
+    {
+      "epoch": 0.9702014958611639,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00029302306689328773,
+      "loss": 2.7251,
+      "step": 19026
+    },
+    {
+      "epoch": 0.9702524893172694,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00029271034846608385,
+      "loss": 2.7003,
+      "step": 19027
+    },
+    {
+      "epoch": 0.9703034827733747,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.000292398128015698,
+      "loss": 2.7014,
+      "step": 19028
+    },
+    {
+      "epoch": 0.9703544762294801,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002920864057347234,
+      "loss": 2.7122,
+      "step": 19029
+    },
+    {
+      "epoch": 0.9704054696855855,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00029177518181544603,
+      "loss": 2.7231,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9704564631416909,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002914644564498444,
+      "loss": 2.6853,
+      "step": 19031
+    },
+    {
+      "epoch": 0.9705074565977962,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002911542298295895,
+      "loss": 2.715,
+      "step": 19032
+    },
+    {
+      "epoch": 0.9705584500539016,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00029084450214604486,
+      "loss": 2.7252,
+      "step": 19033
+    },
+    {
+      "epoch": 0.9706094435100071,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.000290535273590266,
+      "loss": 2.6865,
+      "step": 19034
+    },
+    {
+      "epoch": 0.9706604369661125,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00029022654435300086,
+      "loss": 2.7175,
+      "step": 19035
+    },
+    {
+      "epoch": 0.9707114304222179,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00028991831462468856,
+      "loss": 2.7279,
+      "step": 19036
+    },
+    {
+      "epoch": 0.9707624238783232,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00028961058459546106,
+      "loss": 2.7062,
+      "step": 19037
+    },
+    {
+      "epoch": 0.9708134173344286,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00028930335445514176,
+      "loss": 2.6928,
+      "step": 19038
+    },
+    {
+      "epoch": 0.970864410790534,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0002889966243932456,
+      "loss": 2.714,
+      "step": 19039
+    },
+    {
+      "epoch": 0.9709154042466394,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00028869039459897885,
+      "loss": 2.7229,
+      "step": 19040
+    },
+    {
+      "epoch": 0.9709663977027448,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002883846652612399,
+      "loss": 2.7222,
+      "step": 19041
+    },
+    {
+      "epoch": 0.9710173911588502,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002880794365686176,
+      "loss": 2.7143,
+      "step": 19042
+    },
+    {
+      "epoch": 0.9710683846149556,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00028777470870939236,
+      "loss": 2.7286,
+      "step": 19043
+    },
+    {
+      "epoch": 0.971119378071061,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0002874704818715357,
+      "loss": 2.7498,
+      "step": 19044
+    },
+    {
+      "epoch": 0.9711703715271663,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00028716675624271005,
+      "loss": 2.7107,
+      "step": 19045
+    },
+    {
+      "epoch": 0.9712213649832717,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0002868635320102687,
+      "loss": 2.7214,
+      "step": 19046
+    },
+    {
+      "epoch": 0.9712723584393771,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002865608093612555,
+      "loss": 2.7297,
+      "step": 19047
+    },
+    {
+      "epoch": 0.9713233518954826,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.000286258588482405,
+      "loss": 2.6974,
+      "step": 19048
+    },
+    {
+      "epoch": 0.971374345351588,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002859568695601422,
+      "loss": 2.7074,
+      "step": 19049
+    },
+    {
+      "epoch": 0.9714253388076933,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002856556527805826,
+      "loss": 2.7304,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9714763322637987,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0002853549383295316,
+      "loss": 2.7223,
+      "step": 19051
+    },
+    {
+      "epoch": 0.9715273257199041,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00028505472639248514,
+      "loss": 2.6969,
+      "step": 19052
+    },
+    {
+      "epoch": 0.9715783191760095,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00028475501715462915,
+      "loss": 2.7158,
+      "step": 19053
+    },
+    {
+      "epoch": 0.9716293126321148,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002844558108008392,
+      "loss": 2.6792,
+      "step": 19054
+    },
+    {
+      "epoch": 0.9716803060882203,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00028415710751568087,
+      "loss": 2.7326,
+      "step": 19055
+    },
+    {
+      "epoch": 0.9717312995443257,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002838589074834093,
+      "loss": 2.7445,
+      "step": 19056
+    },
+    {
+      "epoch": 0.9717822930004311,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00028356121088796913,
+      "loss": 2.6953,
+      "step": 19057
+    },
+    {
+      "epoch": 0.9718332864565364,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00028326401791299485,
+      "loss": 2.7572,
+      "step": 19058
+    },
+    {
+      "epoch": 0.9718842799126418,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002829673287418099,
+      "loss": 2.7433,
+      "step": 19059
+    },
+    {
+      "epoch": 0.9719352733687472,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002826711435574268,
+      "loss": 2.7248,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9719862668248526,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00028237546254254785,
+      "loss": 2.7113,
+      "step": 19061
+    },
+    {
+      "epoch": 0.972037260280958,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00028208028587956393,
+      "loss": 2.6911,
+      "step": 19062
+    },
+    {
+      "epoch": 0.9720882537370634,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002817856137505546,
+      "loss": 2.7008,
+      "step": 19063
+    },
+    {
+      "epoch": 0.9721392471931688,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0002814914463372886,
+      "loss": 2.7063,
+      "step": 19064
+    },
+    {
+      "epoch": 0.9721902406492742,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002811977838212231,
+      "loss": 2.7011,
+      "step": 19065
+    },
+    {
+      "epoch": 0.9722412341053795,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00028090462638350394,
+      "loss": 2.7404,
+      "step": 19066
+    },
+    {
+      "epoch": 0.9722922275614849,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0002806119742049653,
+      "loss": 2.7148,
+      "step": 19067
+    },
+    {
+      "epoch": 0.9723432210175903,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00028031982746612964,
+      "loss": 2.7206,
+      "step": 19068
+    },
+    {
+      "epoch": 0.9723942144736958,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00028002818634720805,
+      "loss": 2.719,
+      "step": 19069
+    },
+    {
+      "epoch": 0.9724452079298012,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002797370510280993,
+      "loss": 2.7128,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9724962013859065,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002794464216883903,
+      "loss": 2.7269,
+      "step": 19071
+    },
+    {
+      "epoch": 0.9725471948420119,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0002791562985073555,
+      "loss": 2.7131,
+      "step": 19072
+    },
+    {
+      "epoch": 0.9725981882981173,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0002788666816639579,
+      "loss": 2.7368,
+      "step": 19073
+    },
+    {
+      "epoch": 0.9726491817542227,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00027857757133684753,
+      "loss": 2.7191,
+      "step": 19074
+    },
+    {
+      "epoch": 0.972700175210328,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0002782889677043622,
+      "loss": 2.7317,
+      "step": 19075
+    },
+    {
+      "epoch": 0.9727511686664335,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00027800087094452693,
+      "loss": 2.708,
+      "step": 19076
+    },
+    {
+      "epoch": 0.9728021621225389,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002777132812350546,
+      "loss": 2.7128,
+      "step": 19077
+    },
+    {
+      "epoch": 0.9728531555786443,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00027742619875334505,
+      "loss": 2.708,
+      "step": 19078
+    },
+    {
+      "epoch": 0.9729041490347496,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00027713962367648484,
+      "loss": 2.7299,
+      "step": 19079
+    },
+    {
+      "epoch": 0.972955142490855,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00027685355618124773,
+      "loss": 2.713,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9730061359469604,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002765679964440951,
+      "loss": 2.7473,
+      "step": 19081
+    },
+    {
+      "epoch": 0.9730571294030658,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00027628294464117424,
+      "loss": 2.7247,
+      "step": 19082
+    },
+    {
+      "epoch": 0.9731081228591713,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002759984009483195,
+      "loss": 2.7035,
+      "step": 19083
+    },
+    {
+      "epoch": 0.9731591163152766,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00027571436554105146,
+      "loss": 2.6928,
+      "step": 19084
+    },
+    {
+      "epoch": 0.973210109771382,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00027543083859457774,
+      "loss": 2.6922,
+      "step": 19085
+    },
+    {
+      "epoch": 0.9732611032274874,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002751478202837919,
+      "loss": 2.7052,
+      "step": 19086
+    },
+    {
+      "epoch": 0.9733120966835928,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00027486531078327385,
+      "loss": 2.7223,
+      "step": 19087
+    },
+    {
+      "epoch": 0.9733630901396981,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00027458331026728957,
+      "loss": 2.7078,
+      "step": 19088
+    },
+    {
+      "epoch": 0.9734140835958035,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00027430181890979134,
+      "loss": 2.7174,
+      "step": 19089
+    },
+    {
+      "epoch": 0.973465077051909,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.000274020836884417,
+      "loss": 2.7097,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9735160705080144,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002737403643644905,
+      "loss": 2.7082,
+      "step": 19091
+    },
+    {
+      "epoch": 0.9735670639641197,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002734604015230213,
+      "loss": 2.7427,
+      "step": 19092
+    },
+    {
+      "epoch": 0.9736180574202251,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00027318094853270446,
+      "loss": 2.7164,
+      "step": 19093
+    },
+    {
+      "epoch": 0.9736690508763305,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00027290200556592093,
+      "loss": 2.7196,
+      "step": 19094
+    },
+    {
+      "epoch": 0.9737200443324359,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00027262357279473644,
+      "loss": 2.7272,
+      "step": 19095
+    },
+    {
+      "epoch": 0.9737710377885412,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002723456503909022,
+      "loss": 2.7162,
+      "step": 19096
+    },
+    {
+      "epoch": 0.9738220312446467,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002720682385258552,
+      "loss": 2.7074,
+      "step": 19097
+    },
+    {
+      "epoch": 0.9738730247007521,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002717913373707167,
+      "loss": 2.6942,
+      "step": 19098
+    },
+    {
+      "epoch": 0.9739240181568575,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002715149470962933,
+      "loss": 2.7335,
+      "step": 19099
+    },
+    {
+      "epoch": 0.9739750116129628,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00027123906787307646,
+      "loss": 2.7193,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9740260050690682,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002709636998712423,
+      "loss": 2.6827,
+      "step": 19101
+    },
+    {
+      "epoch": 0.9740769985251736,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00027068884326065146,
+      "loss": 2.7023,
+      "step": 19102
+    },
+    {
+      "epoch": 0.974127991981279,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00027041449821084955,
+      "loss": 2.7135,
+      "step": 19103
+    },
+    {
+      "epoch": 0.9741789854373845,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002701406648910662,
+      "loss": 2.6951,
+      "step": 19104
+    },
+    {
+      "epoch": 0.9742299788934898,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00026986734347021583,
+      "loss": 2.7181,
+      "step": 19105
+    },
+    {
+      "epoch": 0.9742809723495952,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002695945341168966,
+      "loss": 2.6903,
+      "step": 19106
+    },
+    {
+      "epoch": 0.9743319658057006,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00026932223699939137,
+      "loss": 2.6964,
+      "step": 19107
+    },
+    {
+      "epoch": 0.974382959261806,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00026905045228566596,
+      "loss": 2.7024,
+      "step": 19108
+    },
+    {
+      "epoch": 0.9744339527179113,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002687791801433715,
+      "loss": 2.6955,
+      "step": 19109
+    },
+    {
+      "epoch": 0.9744849461740167,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00026850842073984193,
+      "loss": 2.7027,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9745359396301222,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00026823817424209543,
+      "loss": 2.7027,
+      "step": 19111
+    },
+    {
+      "epoch": 0.9745869330862276,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002679684408168337,
+      "loss": 2.6981,
+      "step": 19112
+    },
+    {
+      "epoch": 0.9746379265423329,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002676992206304415,
+      "loss": 2.7071,
+      "step": 19113
+    },
+    {
+      "epoch": 0.9746889199984383,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00026743051384898786,
+      "loss": 2.7053,
+      "step": 19114
+    },
+    {
+      "epoch": 0.9747399134545437,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002671623206382242,
+      "loss": 2.7426,
+      "step": 19115
+    },
+    {
+      "epoch": 0.9747909069106491,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00026689464116358587,
+      "loss": 2.7196,
+      "step": 19116
+    },
+    {
+      "epoch": 0.9748419003667544,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00026662747559019057,
+      "loss": 2.6976,
+      "step": 19117
+    },
+    {
+      "epoch": 0.9748928938228599,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00026636082408284004,
+      "loss": 2.7254,
+      "step": 19118
+    },
+    {
+      "epoch": 0.9749438872789653,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.000266094686806018,
+      "loss": 2.7391,
+      "step": 19119
+    },
+    {
+      "epoch": 0.9749948807350707,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002658290639238913,
+      "loss": 2.716,
+      "step": 19120
+    },
+    {
+      "epoch": 0.975045874191176,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00026556395560030957,
+      "loss": 2.7063,
+      "step": 19121
+    },
+    {
+      "epoch": 0.9750968676472814,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002652993619988049,
+      "loss": 2.7152,
+      "step": 19122
+    },
+    {
+      "epoch": 0.9751478611033868,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00026503528328259194,
+      "loss": 2.7179,
+      "step": 19123
+    },
+    {
+      "epoch": 0.9751988545594922,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00026477171961456766,
+      "loss": 2.702,
+      "step": 19124
+    },
+    {
+      "epoch": 0.9752498480155977,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002645086711573112,
+      "loss": 2.7465,
+      "step": 19125
+    },
+    {
+      "epoch": 0.975300841471703,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002642461380730844,
+      "loss": 2.7217,
+      "step": 19126
+    },
+    {
+      "epoch": 0.9753518349278084,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00026398412052383076,
+      "loss": 2.7238,
+      "step": 19127
+    },
+    {
+      "epoch": 0.9754028283839138,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002637226186711759,
+      "loss": 2.6986,
+      "step": 19128
+    },
+    {
+      "epoch": 0.9754538218400192,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00026346163267642713,
+      "loss": 2.7146,
+      "step": 19129
+    },
+    {
+      "epoch": 0.9755048152961245,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00026320116270057384,
+      "loss": 2.7232,
+      "step": 19130
+    },
+    {
+      "epoch": 0.9755558087522299,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000262941208904287,
+      "loss": 2.7367,
+      "step": 19131
+    },
+    {
+      "epoch": 0.9756068022083354,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002626817714479192,
+      "loss": 2.7041,
+      "step": 19132
+    },
+    {
+      "epoch": 0.9756577956644408,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002624228504915041,
+      "loss": 2.7169,
+      "step": 19133
+    },
+    {
+      "epoch": 0.9757087891205461,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0002621644461947578,
+      "loss": 2.6869,
+      "step": 19134
+    },
+    {
+      "epoch": 0.9757597825766515,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0002619065587170768,
+      "loss": 2.7089,
+      "step": 19135
+    },
+    {
+      "epoch": 0.9758107760327569,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00026164918821753897,
+      "loss": 2.7025,
+      "step": 19136
+    },
+    {
+      "epoch": 0.9758617694888623,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00026139233485490346,
+      "loss": 2.7198,
+      "step": 19137
+    },
+    {
+      "epoch": 0.9759127629449676,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002611359987876102,
+      "loss": 2.725,
+      "step": 19138
+    },
+    {
+      "epoch": 0.9759637564010731,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002608801801737803,
+      "loss": 2.701,
+      "step": 19139
+    },
+    {
+      "epoch": 0.9760147498571785,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002606248791712154,
+      "loss": 2.7173,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9760657433132839,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00026037009593739795,
+      "loss": 2.7216,
+      "step": 19141
+    },
+    {
+      "epoch": 0.9761167367693893,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00026011583062949124,
+      "loss": 2.7028,
+      "step": 19142
+    },
+    {
+      "epoch": 0.9761677302254946,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.000259862083404339,
+      "loss": 2.7036,
+      "step": 19143
+    },
+    {
+      "epoch": 0.9762187236816,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00025960885441846495,
+      "loss": 2.706,
+      "step": 19144
+    },
+    {
+      "epoch": 0.9762697171377054,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00025935614382807334,
+      "loss": 2.771,
+      "step": 19145
+    },
+    {
+      "epoch": 0.9763207105938109,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0002591039517890492,
+      "loss": 2.7031,
+      "step": 19146
+    },
+    {
+      "epoch": 0.9763717040499162,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002588522784569571,
+      "loss": 2.7211,
+      "step": 19147
+    },
+    {
+      "epoch": 0.9764226975060216,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00025860112398704157,
+      "loss": 2.7066,
+      "step": 19148
+    },
+    {
+      "epoch": 0.976473690962127,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002583504885342274,
+      "loss": 2.7169,
+      "step": 19149
+    },
+    {
+      "epoch": 0.9765246844182324,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002581003722531195,
+      "loss": 2.7459,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9765756778743377,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002578507752980017,
+      "loss": 2.7249,
+      "step": 19151
+    },
+    {
+      "epoch": 0.9766266713304431,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002576016978228382,
+      "loss": 2.726,
+      "step": 19152
+    },
+    {
+      "epoch": 0.9766776647865486,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002573531399812721,
+      "loss": 2.6997,
+      "step": 19153
+    },
+    {
+      "epoch": 0.976728658242654,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00025710510192662695,
+      "loss": 2.7046,
+      "step": 19154
+    },
+    {
+      "epoch": 0.9767796516987594,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00025685758381190473,
+      "loss": 2.7278,
+      "step": 19155
+    },
+    {
+      "epoch": 0.9768306451548647,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002566105857897871,
+      "loss": 2.7169,
+      "step": 19156
+    },
+    {
+      "epoch": 0.9768816386109701,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002563641080126349,
+      "loss": 2.7513,
+      "step": 19157
+    },
+    {
+      "epoch": 0.9769326320670755,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00025611815063248804,
+      "loss": 2.7192,
+      "step": 19158
+    },
+    {
+      "epoch": 0.9769836255231809,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002558727138010653,
+      "loss": 2.7315,
+      "step": 19159
+    },
+    {
+      "epoch": 0.9770346189792863,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00025562779766976436,
+      "loss": 2.7176,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9770856124353917,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002553834023896619,
+      "loss": 2.7102,
+      "step": 19161
+    },
+    {
+      "epoch": 0.9771366058914971,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002551395281115134,
+      "loss": 2.7145,
+      "step": 19162
+    },
+    {
+      "epoch": 0.9771875993476025,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002548961749857524,
+      "loss": 2.7306,
+      "step": 19163
+    },
+    {
+      "epoch": 0.9772385928037078,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00025465334316249167,
+      "loss": 2.7426,
+      "step": 19164
+    },
+    {
+      "epoch": 0.9772895862598132,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00025441103279152195,
+      "loss": 2.7155,
+      "step": 19165
+    },
+    {
+      "epoch": 0.9773405797159186,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00025416924402231234,
+      "loss": 2.7061,
+      "step": 19166
+    },
+    {
+      "epoch": 0.9773915731720241,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0002539279770040105,
+      "loss": 2.6875,
+      "step": 19167
+    },
+    {
+      "epoch": 0.9774425666281295,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00025368723188544184,
+      "loss": 2.7236,
+      "step": 19168
+    },
+    {
+      "epoch": 0.9774935600842348,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00025344700881511006,
+      "loss": 2.7178,
+      "step": 19169
+    },
+    {
+      "epoch": 0.9775445535403402,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0002532073079411971,
+      "loss": 2.7051,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9775955469964456,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00025296812941156226,
+      "loss": 2.7096,
+      "step": 19171
+    },
+    {
+      "epoch": 0.977646540452551,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0002527294733737429,
+      "loss": 2.6969,
+      "step": 19172
+    },
+    {
+      "epoch": 0.9776975339086563,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00025249133997495404,
+      "loss": 2.7436,
+      "step": 19173
+    },
+    {
+      "epoch": 0.9777485273647618,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002522537293620885,
+      "loss": 2.7059,
+      "step": 19174
+    },
+    {
+      "epoch": 0.9777995208208672,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0002520166416817162,
+      "loss": 2.7201,
+      "step": 19175
+    },
+    {
+      "epoch": 0.9778505142769726,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00025178007708008494,
+      "loss": 2.6779,
+      "step": 19176
+    },
+    {
+      "epoch": 0.9779015077330779,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00025154403570311936,
+      "loss": 2.719,
+      "step": 19177
+    },
+    {
+      "epoch": 0.9779525011891833,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002513085176964219,
+      "loss": 2.7468,
+      "step": 19178
+    },
+    {
+      "epoch": 0.9780034946452887,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000251073523205272,
+      "loss": 2.7386,
+      "step": 19179
+    },
+    {
+      "epoch": 0.9780544881013941,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0002508390523746259,
+      "loss": 2.6936,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9781054815574995,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00025060510534911684,
+      "loss": 2.6966,
+      "step": 19181
+    },
+    {
+      "epoch": 0.9781564750136049,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0002503716822730553,
+      "loss": 2.7042,
+      "step": 19182
+    },
+    {
+      "epoch": 0.9782074684697103,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0002501387832904284,
+      "loss": 2.7132,
+      "step": 19183
+    },
+    {
+      "epoch": 0.9782584619258157,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002499064085448997,
+      "loss": 2.7165,
+      "step": 19184
+    },
+    {
+      "epoch": 0.978309455381921,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002496745581798098,
+      "loss": 2.6998,
+      "step": 19185
+    },
+    {
+      "epoch": 0.9783604488380264,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0002494432323381758,
+      "loss": 2.7158,
+      "step": 19186
+    },
+    {
+      "epoch": 0.9784114422941318,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0002492124311626907,
+      "loss": 2.7073,
+      "step": 19187
+    },
+    {
+      "epoch": 0.9784624357502373,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002489821547957246,
+      "loss": 2.7348,
+      "step": 19188
+    },
+    {
+      "epoch": 0.9785134292063427,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002487524033793233,
+      "loss": 2.7179,
+      "step": 19189
+    },
+    {
+      "epoch": 0.978564422662448,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00024852317705520926,
+      "loss": 2.6703,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9786154161185534,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00024829447596478064,
+      "loss": 2.7593,
+      "step": 19191
+    },
+    {
+      "epoch": 0.9786664095746588,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002480663002491117,
+      "loss": 2.7275,
+      "step": 19192
+    },
+    {
+      "epoch": 0.9787174030307642,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00024783865004895293,
+      "loss": 2.7303,
+      "step": 19193
+    },
+    {
+      "epoch": 0.9787683964868695,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002476115255047302,
+      "loss": 2.6976,
+      "step": 19194
+    },
+    {
+      "epoch": 0.978819389942975,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002473849267565454,
+      "loss": 2.7357,
+      "step": 19195
+    },
+    {
+      "epoch": 0.9788703833990804,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0002471588539441761,
+      "loss": 2.7335,
+      "step": 19196
+    },
+    {
+      "epoch": 0.9789213768551858,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000246933307207075,
+      "loss": 2.7162,
+      "step": 19197
+    },
+    {
+      "epoch": 0.9789723703112911,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002467082866843712,
+      "loss": 2.7096,
+      "step": 19198
+    },
+    {
+      "epoch": 0.9790233637673965,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00024648379251486843,
+      "loss": 2.7086,
+      "step": 19199
+    },
+    {
+      "epoch": 0.9790743572235019,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00024625982483704586,
+      "loss": 2.7084,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9791253506796073,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002460363837890582,
+      "loss": 2.7181,
+      "step": 19201
+    },
+    {
+      "epoch": 0.9791763441357128,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000245813469508735,
+      "loss": 2.7124,
+      "step": 19202
+    },
+    {
+      "epoch": 0.9792273375918181,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002455910821335809,
+      "loss": 2.7037,
+      "step": 19203
+    },
+    {
+      "epoch": 0.9792783310479235,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002453692218007758,
+      "loss": 2.6909,
+      "step": 19204
+    },
+    {
+      "epoch": 0.9793293245040289,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00024514788864717405,
+      "loss": 2.7288,
+      "step": 19205
+    },
+    {
+      "epoch": 0.9793803179601343,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0002449270828093054,
+      "loss": 2.6918,
+      "step": 19206
+    },
+    {
+      "epoch": 0.9794313114162396,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00024470680442337386,
+      "loss": 2.7399,
+      "step": 19207
+    },
+    {
+      "epoch": 0.979482304872345,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00024448705362525803,
+      "loss": 2.731,
+      "step": 19208
+    },
+    {
+      "epoch": 0.9795332983284505,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002442678305505114,
+      "loss": 2.7171,
+      "step": 19209
+    },
+    {
+      "epoch": 0.9795842917845559,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00024404913533436186,
+      "loss": 2.7061,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9796352852406612,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00024383096811171138,
+      "loss": 2.7308,
+      "step": 19211
+    },
+    {
+      "epoch": 0.9796862786967666,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002436133290171366,
+      "loss": 2.7541,
+      "step": 19212
+    },
+    {
+      "epoch": 0.979737272152872,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002433962181848881,
+      "loss": 2.7016,
+      "step": 19213
+    },
+    {
+      "epoch": 0.9797882656089774,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00024317963574889106,
+      "loss": 2.6998,
+      "step": 19214
+    },
+    {
+      "epoch": 0.9798392590650827,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00024296358184274408,
+      "loss": 2.708,
+      "step": 19215
+    },
+    {
+      "epoch": 0.9798902525211882,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0002427480565997205,
+      "loss": 2.6815,
+      "step": 19216
+    },
+    {
+      "epoch": 0.9799412459772936,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00024253306015276644,
+      "loss": 2.7455,
+      "step": 19217
+    },
+    {
+      "epoch": 0.979992239433399,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00024231859263450315,
+      "loss": 2.6965,
+      "step": 19218
+    },
+    {
+      "epoch": 0.9800432328895043,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00024210465417722455,
+      "loss": 2.7141,
+      "step": 19219
+    },
+    {
+      "epoch": 0.9800942263456097,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00024189124491289893,
+      "loss": 2.7122,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9801452198017151,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00024167836497316755,
+      "loss": 2.6972,
+      "step": 19221
+    },
+    {
+      "epoch": 0.9801962132578205,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00024146601448934592,
+      "loss": 2.7053,
+      "step": 19222
+    },
+    {
+      "epoch": 0.980247206713926,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00024125419359242198,
+      "loss": 2.7339,
+      "step": 19223
+    },
+    {
+      "epoch": 0.9802982001700313,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00024104290241305768,
+      "loss": 2.7251,
+      "step": 19224
+    },
+    {
+      "epoch": 0.9803491936261367,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00024083214108158805,
+      "loss": 2.7148,
+      "step": 19225
+    },
+    {
+      "epoch": 0.9804001870822421,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00024062190972802137,
+      "loss": 2.7115,
+      "step": 19226
+    },
+    {
+      "epoch": 0.9804511805383475,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00024041220848203883,
+      "loss": 2.6919,
+      "step": 19227
+    },
+    {
+      "epoch": 0.9805021739944528,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002402030374729946,
+      "loss": 2.7026,
+      "step": 19228
+    },
+    {
+      "epoch": 0.9805531674505582,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00023999439682991598,
+      "loss": 2.7185,
+      "step": 19229
+    },
+    {
+      "epoch": 0.9806041609066637,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002397862866815029,
+      "loss": 2.7173,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9806551543627691,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00023957870715612836,
+      "loss": 2.7167,
+      "step": 19231
+    },
+    {
+      "epoch": 0.9807061478188744,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002393716583818376,
+      "loss": 2.7383,
+      "step": 19232
+    },
+    {
+      "epoch": 0.9807571412749798,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00023916514048634874,
+      "loss": 2.702,
+      "step": 19233
+    },
+    {
+      "epoch": 0.9808081347310852,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023895915359705254,
+      "loss": 2.713,
+      "step": 19234
+    },
+    {
+      "epoch": 0.9808591281871906,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00023875369784101208,
+      "loss": 2.6927,
+      "step": 19235
+    },
+    {
+      "epoch": 0.9809101216432959,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0002385487733449626,
+      "loss": 2.7009,
+      "step": 19236
+    },
+    {
+      "epoch": 0.9809611150994014,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00023834438023531197,
+      "loss": 2.744,
+      "step": 19237
+    },
+    {
+      "epoch": 0.9810121085555068,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00023814051863814002,
+      "loss": 2.7324,
+      "step": 19238
+    },
+    {
+      "epoch": 0.9810631020116122,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023793718867919891,
+      "loss": 2.7063,
+      "step": 19239
+    },
+    {
+      "epoch": 0.9811140954677176,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00023773439048391283,
+      "loss": 2.7234,
+      "step": 19240
+    },
+    {
+      "epoch": 0.9811650889238229,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00023753212417737776,
+      "loss": 2.7163,
+      "step": 19241
+    },
+    {
+      "epoch": 0.9812160823799283,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.00023733038988436167,
+      "loss": 2.7061,
+      "step": 19242
+    },
+    {
+      "epoch": 0.9812670758360337,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00023712918772930463,
+      "loss": 2.6927,
+      "step": 19243
+    },
+    {
+      "epoch": 0.9813180692921392,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00023692851783631817,
+      "loss": 2.6984,
+      "step": 19244
+    },
+    {
+      "epoch": 0.9813690627482445,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00023672838032918556,
+      "loss": 2.732,
+      "step": 19245
+    },
+    {
+      "epoch": 0.9814200562043499,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00023652877533136151,
+      "loss": 2.714,
+      "step": 19246
+    },
+    {
+      "epoch": 0.9814710496604553,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00023632970296597253,
+      "loss": 2.7176,
+      "step": 19247
+    },
+    {
+      "epoch": 0.9815220431165607,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00023613116335581647,
+      "loss": 2.7118,
+      "step": 19248
+    },
+    {
+      "epoch": 0.981573036572666,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0002359331566233626,
+      "loss": 2.7267,
+      "step": 19249
+    },
+    {
+      "epoch": 0.9816240300287714,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00023573568289075126,
+      "loss": 2.7351,
+      "step": 19250
+    },
+    {
+      "epoch": 0.9816750234848769,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00023553874227979425,
+      "loss": 2.7144,
+      "step": 19251
+    },
+    {
+      "epoch": 0.9817260169409823,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002353423349119746,
+      "loss": 2.6874,
+      "step": 19252
+    },
+    {
+      "epoch": 0.9817770103970876,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00023514646090844597,
+      "loss": 2.7311,
+      "step": 19253
+    },
+    {
+      "epoch": 0.981828003853193,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002349511203900333,
+      "loss": 2.7343,
+      "step": 19254
+    },
+    {
+      "epoch": 0.9818789973092984,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00023475631347723267,
+      "loss": 2.7152,
+      "step": 19255
+    },
+    {
+      "epoch": 0.9819299907654038,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00023456204029021058,
+      "loss": 2.7186,
+      "step": 19256
+    },
+    {
+      "epoch": 0.9819809842215091,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00023436830094880452,
+      "loss": 2.6964,
+      "step": 19257
+    },
+    {
+      "epoch": 0.9820319776776146,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00023417509557252267,
+      "loss": 2.7266,
+      "step": 19258
+    },
+    {
+      "epoch": 0.98208297113372,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000233982424280544,
+      "loss": 2.6875,
+      "step": 19259
+    },
+    {
+      "epoch": 0.9821339645898254,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002337902871917176,
+      "loss": 2.6778,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9821849580459308,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00023359868442456323,
+      "loss": 2.6899,
+      "step": 19261
+    },
+    {
+      "epoch": 0.9822359515020361,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00023340761609727116,
+      "loss": 2.7101,
+      "step": 19262
+    },
+    {
+      "epoch": 0.9822869449581415,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00023321708232770216,
+      "loss": 2.7096,
+      "step": 19263
+    },
+    {
+      "epoch": 0.9823379384142469,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00023302708323338683,
+      "loss": 2.7268,
+      "step": 19264
+    },
+    {
+      "epoch": 0.9823889318703524,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00023283761893152613,
+      "loss": 2.719,
+      "step": 19265
+    },
+    {
+      "epoch": 0.9824399253264577,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00023264868953899127,
+      "loss": 2.6942,
+      "step": 19266
+    },
+    {
+      "epoch": 0.9824909187825631,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002324602951723234,
+      "loss": 2.706,
+      "step": 19267
+    },
+    {
+      "epoch": 0.9825419122386685,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002322724359477335,
+      "loss": 2.6983,
+      "step": 19268
+    },
+    {
+      "epoch": 0.9825929056947739,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00023208511198110276,
+      "loss": 2.7401,
+      "step": 19269
+    },
+    {
+      "epoch": 0.9826438991508792,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00023189832338798175,
+      "loss": 2.7131,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9826948926069846,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00023171207028359128,
+      "loss": 2.7075,
+      "step": 19271
+    },
+    {
+      "epoch": 0.9827458860630901,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00023152635278282158,
+      "loss": 2.7227,
+      "step": 19272
+    },
+    {
+      "epoch": 0.9827968795191955,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0002313411710002326,
+      "loss": 2.7255,
+      "step": 19273
+    },
+    {
+      "epoch": 0.9828478729753009,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.00023115652505005353,
+      "loss": 2.7009,
+      "step": 19274
+    },
+    {
+      "epoch": 0.9828988664314062,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00023097241504618348,
+      "loss": 2.7082,
+      "step": 19275
+    },
+    {
+      "epoch": 0.9829498598875116,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00023078884110219077,
+      "loss": 2.6952,
+      "step": 19276
+    },
+    {
+      "epoch": 0.983000853343617,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00023060580333131295,
+      "loss": 2.7054,
+      "step": 19277
+    },
+    {
+      "epoch": 0.9830518467997224,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002304233018464567,
+      "loss": 2.7226,
+      "step": 19278
+    },
+    {
+      "epoch": 0.9831028402558278,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0002302413367601985,
+      "loss": 2.7081,
+      "step": 19279
+    },
+    {
+      "epoch": 0.9831538337119332,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00023005990818478347,
+      "loss": 2.7065,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9832048271680386,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00022987901623212575,
+      "loss": 2.7116,
+      "step": 19281
+    },
+    {
+      "epoch": 0.983255820624144,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00022969866101380852,
+      "loss": 2.7085,
+      "step": 19282
+    },
+    {
+      "epoch": 0.9833068140802493,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00022951884264108406,
+      "loss": 2.7075,
+      "step": 19283
+    },
+    {
+      "epoch": 0.9833578075363547,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00022933956122487333,
+      "loss": 2.7128,
+      "step": 19284
+    },
+    {
+      "epoch": 0.9834088009924601,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00022916081687576623,
+      "loss": 2.7287,
+      "step": 19285
+    },
+    {
+      "epoch": 0.9834597944485656,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002289826097040211,
+      "loss": 2.7137,
+      "step": 19286
+    },
+    {
+      "epoch": 0.983510787904671,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00022880493981956512,
+      "loss": 2.7463,
+      "step": 19287
+    },
+    {
+      "epoch": 0.9835617813607763,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00022862780733199416,
+      "loss": 2.721,
+      "step": 19288
+    },
+    {
+      "epoch": 0.9836127748168817,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00022845121235057228,
+      "loss": 2.7436,
+      "step": 19289
+    },
+    {
+      "epoch": 0.9836637682729871,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00022827515498423205,
+      "loss": 2.7067,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9837147617290924,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002280996353415746,
+      "loss": 2.7213,
+      "step": 19291
+    },
+    {
+      "epoch": 0.9837657551851978,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0002279246535308694,
+      "loss": 2.6879,
+      "step": 19292
+    },
+    {
+      "epoch": 0.9838167486413033,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0002277502096600539,
+      "loss": 2.7399,
+      "step": 19293
+    },
+    {
+      "epoch": 0.9838677420974087,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00022757630383673358,
+      "loss": 2.7174,
+      "step": 19294
+    },
+    {
+      "epoch": 0.9839187355535141,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00022740293616818285,
+      "loss": 2.7273,
+      "step": 19295
+    },
+    {
+      "epoch": 0.9839697290096194,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0002272301067613432,
+      "loss": 2.7422,
+      "step": 19296
+    },
+    {
+      "epoch": 0.9840207224657248,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00022705781572282462,
+      "loss": 2.7021,
+      "step": 19297
+    },
+    {
+      "epoch": 0.9840717159218302,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00022688606315890482,
+      "loss": 2.705,
+      "step": 19298
+    },
+    {
+      "epoch": 0.9841227093779356,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022671484917552936,
+      "loss": 2.6978,
+      "step": 19299
+    },
+    {
+      "epoch": 0.984173702834041,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002265441738783118,
+      "loss": 2.7014,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9842246962901464,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002263740373725332,
+      "loss": 2.7319,
+      "step": 19301
+    },
+    {
+      "epoch": 0.9842756897462518,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002262044397631421,
+      "loss": 2.7488,
+      "step": 19302
+    },
+    {
+      "epoch": 0.9843266832023572,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00022603538115475516,
+      "loss": 2.7642,
+      "step": 19303
+    },
+    {
+      "epoch": 0.9843776766584625,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00022586686165165592,
+      "loss": 2.7103,
+      "step": 19304
+    },
+    {
+      "epoch": 0.9844286701145679,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00022569888135779583,
+      "loss": 2.6863,
+      "step": 19305
+    },
+    {
+      "epoch": 0.9844796635706733,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00022553144037679344,
+      "loss": 2.7528,
+      "step": 19306
+    },
+    {
+      "epoch": 0.9845306570267788,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00022536453881193496,
+      "loss": 2.6879,
+      "step": 19307
+    },
+    {
+      "epoch": 0.9845816504828842,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00022519817676617363,
+      "loss": 2.7069,
+      "step": 19308
+    },
+    {
+      "epoch": 0.9846326439389895,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00022503235434212976,
+      "loss": 2.7057,
+      "step": 19309
+    },
+    {
+      "epoch": 0.9846836373950949,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00022486707164209112,
+      "loss": 2.7153,
+      "step": 19310
+    },
+    {
+      "epoch": 0.9847346308512003,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0002247023287680123,
+      "loss": 2.7149,
+      "step": 19311
+    },
+    {
+      "epoch": 0.9847856243073057,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002245381258215149,
+      "loss": 2.7023,
+      "step": 19312
+    },
+    {
+      "epoch": 0.984836617763411,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00022437446290388775,
+      "loss": 2.7296,
+      "step": 19313
+    },
+    {
+      "epoch": 0.9848876112195165,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00022421134011608605,
+      "loss": 2.7238,
+      "step": 19314
+    },
+    {
+      "epoch": 0.9849386046756219,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00022404875755873254,
+      "loss": 2.7011,
+      "step": 19315
+    },
+    {
+      "epoch": 0.9849895981317273,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00022388671533211607,
+      "loss": 2.7004,
+      "step": 19316
+    },
+    {
+      "epoch": 0.9850405915878326,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002237252135361924,
+      "loss": 2.7047,
+      "step": 19317
+    },
+    {
+      "epoch": 0.985091585043938,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00022356425227058402,
+      "loss": 2.7165,
+      "step": 19318
+    },
+    {
+      "epoch": 0.9851425785000434,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00022340383163457994,
+      "loss": 2.7384,
+      "step": 19319
+    },
+    {
+      "epoch": 0.9851935719561488,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0002232439517271356,
+      "loss": 2.7357,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9852445654122542,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00022308461264687311,
+      "loss": 2.699,
+      "step": 19321
+    },
+    {
+      "epoch": 0.9852955588683596,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00022292581449208048,
+      "loss": 2.7179,
+      "step": 19322
+    },
+    {
+      "epoch": 0.985346552324465,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00022276755736071286,
+      "loss": 2.7079,
+      "step": 19323
+    },
+    {
+      "epoch": 0.9853975457805704,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.000222609841350391,
+      "loss": 2.732,
+      "step": 19324
+    },
+    {
+      "epoch": 0.9854485392366757,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00022245266655840192,
+      "loss": 2.6851,
+      "step": 19325
+    },
+    {
+      "epoch": 0.9854995326927811,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.000222296033081699,
+      "loss": 2.7082,
+      "step": 19326
+    },
+    {
+      "epoch": 0.9855505261488865,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00022213994101690166,
+      "loss": 2.6826,
+      "step": 19327
+    },
+    {
+      "epoch": 0.985601519604992,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00022198439046029533,
+      "loss": 2.7159,
+      "step": 19328
+    },
+    {
+      "epoch": 0.9856525130610974,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00022182938150783156,
+      "loss": 2.7264,
+      "step": 19329
+    },
+    {
+      "epoch": 0.9857035065172027,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022167491425512733,
+      "loss": 2.7047,
+      "step": 19330
+    },
+    {
+      "epoch": 0.9857544999733081,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00022152098879746614,
+      "loss": 2.7006,
+      "step": 19331
+    },
+    {
+      "epoch": 0.9858054934294135,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002213676052297968,
+      "loss": 2.7025,
+      "step": 19332
+    },
+    {
+      "epoch": 0.9858564868855189,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00022121476364673397,
+      "loss": 2.6787,
+      "step": 19333
+    },
+    {
+      "epoch": 0.9859074803416242,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002210624641425579,
+      "loss": 2.7204,
+      "step": 19334
+    },
+    {
+      "epoch": 0.9859584737977297,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00022091070681121486,
+      "loss": 2.7334,
+      "step": 19335
+    },
+    {
+      "epoch": 0.9860094672538351,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00022075949174631617,
+      "loss": 2.707,
+      "step": 19336
+    },
+    {
+      "epoch": 0.9860604607099405,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00022060881904113904,
+      "loss": 2.7217,
+      "step": 19337
+    },
+    {
+      "epoch": 0.9861114541660458,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0002204586887886259,
+      "loss": 2.736,
+      "step": 19338
+    },
+    {
+      "epoch": 0.9861624476221512,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00022030910108138463,
+      "loss": 2.7278,
+      "step": 19339
+    },
+    {
+      "epoch": 0.9862134410782566,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00022016005601168848,
+      "loss": 2.7387,
+      "step": 19340
+    },
+    {
+      "epoch": 0.986264434534362,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002200115536714758,
+      "loss": 2.7211,
+      "step": 19341
+    },
+    {
+      "epoch": 0.9863154279904675,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00021986359415235036,
+      "loss": 2.7117,
+      "step": 19342
+    },
+    {
+      "epoch": 0.9863664214465728,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021971617754558114,
+      "loss": 2.7244,
+      "step": 19343
+    },
+    {
+      "epoch": 0.9864174149026782,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.000219569303942102,
+      "loss": 2.733,
+      "step": 19344
+    },
+    {
+      "epoch": 0.9864684083587836,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00021942297343251205,
+      "loss": 2.7035,
+      "step": 19345
+    },
+    {
+      "epoch": 0.986519401814889,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00021927718610707515,
+      "loss": 2.7138,
+      "step": 19346
+    },
+    {
+      "epoch": 0.9865703952709943,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00021913194205572036,
+      "loss": 2.7327,
+      "step": 19347
+    },
+    {
+      "epoch": 0.9866213887270997,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00021898724136804146,
+      "loss": 2.7101,
+      "step": 19348
+    },
+    {
+      "epoch": 0.9866723821832052,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00021884308413329718,
+      "loss": 2.7013,
+      "step": 19349
+    },
+    {
+      "epoch": 0.9867233756393106,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00021869947044041074,
+      "loss": 2.7317,
+      "step": 19350
+    },
+    {
+      "epoch": 0.9867743690954159,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.00021855640037797056,
+      "loss": 2.6981,
+      "step": 19351
+    },
+    {
+      "epoch": 0.9868253625515213,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00021841387403422933,
+      "loss": 2.7046,
+      "step": 19352
+    },
+    {
+      "epoch": 0.9868763560076267,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00021827189149710443,
+      "loss": 2.686,
+      "step": 19353
+    },
+    {
+      "epoch": 0.9869273494637321,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00021813045285417785,
+      "loss": 2.7179,
+      "step": 19354
+    },
+    {
+      "epoch": 0.9869783429198374,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00021798955819269603,
+      "loss": 2.7227,
+      "step": 19355
+    },
+    {
+      "epoch": 0.9870293363759429,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00021784920759956988,
+      "loss": 2.7183,
+      "step": 19356
+    },
+    {
+      "epoch": 0.9870803298320483,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00021770940116137464,
+      "loss": 2.7136,
+      "step": 19357
+    },
+    {
+      "epoch": 0.9871313232881537,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00021757013896435006,
+      "loss": 2.7213,
+      "step": 19358
+    },
+    {
+      "epoch": 0.987182316744259,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00021743142109439996,
+      "loss": 2.7201,
+      "step": 19359
+    },
+    {
+      "epoch": 0.9872333102003644,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00021729324763709263,
+      "loss": 2.7048,
+      "step": 19360
+    },
+    {
+      "epoch": 0.9872843036564698,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00021715561867766028,
+      "loss": 2.7041,
+      "step": 19361
+    },
+    {
+      "epoch": 0.9873352971125752,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00021701853430099944,
+      "loss": 2.7265,
+      "step": 19362
+    },
+    {
+      "epoch": 0.9873862905686807,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002168819945916704,
+      "loss": 2.6971,
+      "step": 19363
+    },
+    {
+      "epoch": 0.987437284024786,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00021674599963389808,
+      "loss": 2.732,
+      "step": 19364
+    },
+    {
+      "epoch": 0.9874882774808914,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00021661054951157085,
+      "loss": 2.7494,
+      "step": 19365
+    },
+    {
+      "epoch": 0.9875392709369968,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002164756443082411,
+      "loss": 2.7282,
+      "step": 19366
+    },
+    {
+      "epoch": 0.9875902643931022,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00021634128410712516,
+      "loss": 2.727,
+      "step": 19367
+    },
+    {
+      "epoch": 0.9876412578492075,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021620746899110314,
+      "loss": 2.7153,
+      "step": 19368
+    },
+    {
+      "epoch": 0.9876922513053129,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000216074199042719,
+      "loss": 2.7111,
+      "step": 19369
+    },
+    {
+      "epoch": 0.9877432447614184,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00021594147434418027,
+      "loss": 2.6727,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9877942382175238,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002158092949773582,
+      "loss": 2.7087,
+      "step": 19371
+    },
+    {
+      "epoch": 0.9878452316736291,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0002156776610237877,
+      "loss": 2.721,
+      "step": 19372
+    },
+    {
+      "epoch": 0.9878962251297345,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00021554657256466723,
+      "loss": 2.7151,
+      "step": 19373
+    },
+    {
+      "epoch": 0.9879472185858399,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00021541602968085878,
+      "loss": 2.6991,
+      "step": 19374
+    },
+    {
+      "epoch": 0.9879982120419453,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002152860324528876,
+      "loss": 2.7273,
+      "step": 19375
+    },
+    {
+      "epoch": 0.9880492054980506,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00021515658096094264,
+      "loss": 2.7289,
+      "step": 19376
+    },
+    {
+      "epoch": 0.9881001989541561,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00021502767528487604,
+      "loss": 2.7029,
+      "step": 19377
+    },
+    {
+      "epoch": 0.9881511924102615,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00021489931550420336,
+      "loss": 2.7264,
+      "step": 19378
+    },
+    {
+      "epoch": 0.9882021858663669,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00021477150169810322,
+      "loss": 2.6992,
+      "step": 19379
+    },
+    {
+      "epoch": 0.9882531793224723,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00021464423394541764,
+      "loss": 2.724,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9883041727785776,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000214517512324652,
+      "loss": 2.7103,
+      "step": 19381
+    },
+    {
+      "epoch": 0.988355166234683,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002143913369139743,
+      "loss": 2.7193,
+      "step": 19382
+    },
+    {
+      "epoch": 0.9884061596907884,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00021426570779121588,
+      "loss": 2.7121,
+      "step": 19383
+    },
+    {
+      "epoch": 0.9884571531468939,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00021414062503387118,
+      "loss": 2.7307,
+      "step": 19384
+    },
+    {
+      "epoch": 0.9885081466029992,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00021401608871909759,
+      "loss": 2.7087,
+      "step": 19385
+    },
+    {
+      "epoch": 0.9885591400591046,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00021389209892371523,
+      "loss": 2.7001,
+      "step": 19386
+    },
+    {
+      "epoch": 0.98861013351521,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00021376865572420722,
+      "loss": 2.7407,
+      "step": 19387
+    },
+    {
+      "epoch": 0.9886611269713154,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00021364575919671967,
+      "loss": 2.7191,
+      "step": 19388
+    },
+    {
+      "epoch": 0.9887121204274207,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00021352340941706123,
+      "loss": 2.7044,
+      "step": 19389
+    },
+    {
+      "epoch": 0.9887631138835261,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00021340160646070354,
+      "loss": 2.6959,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9888141073396316,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00021328035040278057,
+      "loss": 2.7068,
+      "step": 19391
+    },
+    {
+      "epoch": 0.988865100795737,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00021315964131808925,
+      "loss": 2.7176,
+      "step": 19392
+    },
+    {
+      "epoch": 0.9889160942518423,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.000213039479281089,
+      "loss": 2.7168,
+      "step": 19393
+    },
+    {
+      "epoch": 0.9889670877079477,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00021291986436590177,
+      "loss": 2.7391,
+      "step": 19394
+    },
+    {
+      "epoch": 0.9890180811640531,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00021280079664631217,
+      "loss": 2.7117,
+      "step": 19395
+    },
+    {
+      "epoch": 0.9890690746201585,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00021268227619576703,
+      "loss": 2.7197,
+      "step": 19396
+    },
+    {
+      "epoch": 0.9891200680762638,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002125643030873758,
+      "loss": 2.711,
+      "step": 19397
+    },
+    {
+      "epoch": 0.9891710615323693,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00021244687739391014,
+      "loss": 2.7375,
+      "step": 19398
+    },
+    {
+      "epoch": 0.9892220549884747,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00021232999918780404,
+      "loss": 2.7182,
+      "step": 19399
+    },
+    {
+      "epoch": 0.9892730484445801,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00021221366854115413,
+      "loss": 2.7011,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9893240419006855,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002120978855257188,
+      "loss": 2.7616,
+      "step": 19401
+    },
+    {
+      "epoch": 0.9893750353567908,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002119826502129188,
+      "loss": 2.7067,
+      "step": 19402
+    },
+    {
+      "epoch": 0.9894260288128962,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00021186796267383718,
+      "loss": 2.7133,
+      "step": 19403
+    },
+    {
+      "epoch": 0.9894770222690016,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00021175382297921886,
+      "loss": 2.7301,
+      "step": 19404
+    },
+    {
+      "epoch": 0.9895280157251071,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000211640231199471,
+      "loss": 2.6977,
+      "step": 19405
+    },
+    {
+      "epoch": 0.9895790091812124,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002115271874046628,
+      "loss": 2.7226,
+      "step": 19406
+    },
+    {
+      "epoch": 0.9896300026373178,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002114146916645251,
+      "loss": 2.6942,
+      "step": 19407
+    },
+    {
+      "epoch": 0.9896809960934232,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00021130274404845126,
+      "loss": 2.6998,
+      "step": 19408
+    },
+    {
+      "epoch": 0.9897319895495286,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00021119134462549603,
+      "loss": 2.6995,
+      "step": 19409
+    },
+    {
+      "epoch": 0.9897829830056339,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00021108049346437612,
+      "loss": 2.709,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9898339764617393,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00021097019063347013,
+      "loss": 2.7102,
+      "step": 19411
+    },
+    {
+      "epoch": 0.9898849699178448,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00021086043620081838,
+      "loss": 2.7284,
+      "step": 19412
+    },
+    {
+      "epoch": 0.9899359633739502,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00021075123023412296,
+      "loss": 2.6881,
+      "step": 19413
+    },
+    {
+      "epoch": 0.9899869568300556,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00021064257280074762,
+      "loss": 2.7178,
+      "step": 19414
+    },
+    {
+      "epoch": 0.9900379502861609,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00021053446396771764,
+      "loss": 2.7034,
+      "step": 19415
+    },
+    {
+      "epoch": 0.9900889437422663,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00021042690380171996,
+      "loss": 2.7219,
+      "step": 19416
+    },
+    {
+      "epoch": 0.9901399371983717,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002103198923691033,
+      "loss": 2.7306,
+      "step": 19417
+    },
+    {
+      "epoch": 0.990190930654477,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00021021342973587745,
+      "loss": 2.7038,
+      "step": 19418
+    },
+    {
+      "epoch": 0.9902419241105825,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00021010751596771404,
+      "loss": 2.72,
+      "step": 19419
+    },
+    {
+      "epoch": 0.9902929175666879,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00021000215112994595,
+      "loss": 2.7427,
+      "step": 19420
+    },
+    {
+      "epoch": 0.9903439110227933,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00020989733528756753,
+      "loss": 2.7408,
+      "step": 19421
+    },
+    {
+      "epoch": 0.9903949044788987,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020979306850523442,
+      "loss": 2.7254,
+      "step": 19422
+    },
+    {
+      "epoch": 0.990445897935004,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00020968935084726359,
+      "loss": 2.697,
+      "step": 19423
+    },
+    {
+      "epoch": 0.9904968913911094,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00020958618237763347,
+      "loss": 2.673,
+      "step": 19424
+    },
+    {
+      "epoch": 0.9905478848472148,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00020948356315998344,
+      "loss": 2.7225,
+      "step": 19425
+    },
+    {
+      "epoch": 0.9905988783033203,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0002093814932576142,
+      "loss": 2.7245,
+      "step": 19426
+    },
+    {
+      "epoch": 0.9906498717594256,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0002092799727334875,
+      "loss": 2.7084,
+      "step": 19427
+    },
+    {
+      "epoch": 0.990700865215531,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00020917900165022646,
+      "loss": 2.7051,
+      "step": 19428
+    },
+    {
+      "epoch": 0.9907518586716364,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00020907858007011498,
+      "loss": 2.7119,
+      "step": 19429
+    },
+    {
+      "epoch": 0.9908028521277418,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002089787080550982,
+      "loss": 2.7417,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9908538455838471,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00020887938566678223,
+      "loss": 2.7095,
+      "step": 19431
+    },
+    {
+      "epoch": 0.9909048390399525,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00020878061296643408,
+      "loss": 2.6917,
+      "step": 19432
+    },
+    {
+      "epoch": 0.990955832496058,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00020868239001498173,
+      "loss": 2.7266,
+      "step": 19433
+    },
+    {
+      "epoch": 0.9910068259521634,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00020858471687301393,
+      "loss": 2.7102,
+      "step": 19434
+    },
+    {
+      "epoch": 0.9910578194082688,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002084875936007805,
+      "loss": 2.7383,
+      "step": 19435
+    },
+    {
+      "epoch": 0.9911088128643741,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00020839102025819195,
+      "loss": 2.7211,
+      "step": 19436
+    },
+    {
+      "epoch": 0.9911598063204795,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00020829499690481947,
+      "loss": 2.6851,
+      "step": 19437
+    },
+    {
+      "epoch": 0.9912107997765849,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00020819952359989522,
+      "loss": 2.6888,
+      "step": 19438
+    },
+    {
+      "epoch": 0.9912617932326903,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00020810460040231182,
+      "loss": 2.6966,
+      "step": 19439
+    },
+    {
+      "epoch": 0.9913127866887957,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00020801022737062275,
+      "loss": 2.6974,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9913637801449011,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00020791640456304197,
+      "loss": 2.7125,
+      "step": 19441
+    },
+    {
+      "epoch": 0.9914147736010065,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020782313203744418,
+      "loss": 2.701,
+      "step": 19442
+    },
+    {
+      "epoch": 0.9914657670571119,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00020773040985136435,
+      "loss": 2.7063,
+      "step": 19443
+    },
+    {
+      "epoch": 0.9915167605132172,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0002076382380619985,
+      "loss": 2.7035,
+      "step": 19444
+    },
+    {
+      "epoch": 0.9915677539693226,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002075466167262027,
+      "loss": 2.7381,
+      "step": 19445
+    },
+    {
+      "epoch": 0.991618747425428,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002074555459004935,
+      "loss": 2.6887,
+      "step": 19446
+    },
+    {
+      "epoch": 0.9916697408815335,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00020736502564104798,
+      "loss": 2.7099,
+      "step": 19447
+    },
+    {
+      "epoch": 0.9917207343376389,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0002072750560037037,
+      "loss": 2.7057,
+      "step": 19448
+    },
+    {
+      "epoch": 0.9917717277937442,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00020718563704395833,
+      "loss": 2.7135,
+      "step": 19449
+    },
+    {
+      "epoch": 0.9918227212498496,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00020709676881697003,
+      "loss": 2.6632,
+      "step": 19450
+    },
+    {
+      "epoch": 0.991873714705955,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00020700845137755708,
+      "loss": 2.7034,
+      "step": 19451
+    },
+    {
+      "epoch": 0.9919247081620604,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020692068478019832,
+      "loss": 2.7396,
+      "step": 19452
+    },
+    {
+      "epoch": 0.9919757016181657,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00020683346907903252,
+      "loss": 2.7245,
+      "step": 19453
+    },
+    {
+      "epoch": 0.9920266950742712,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00020674680432785866,
+      "loss": 2.714,
+      "step": 19454
+    },
+    {
+      "epoch": 0.9920776885303766,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00020666069058013594,
+      "loss": 2.7128,
+      "step": 19455
+    },
+    {
+      "epoch": 0.992128681986482,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.00020657512788898363,
+      "loss": 2.7082,
+      "step": 19456
+    },
+    {
+      "epoch": 0.9921796754425873,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00020649011630718107,
+      "loss": 2.6877,
+      "step": 19457
+    },
+    {
+      "epoch": 0.9922306688986927,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0002064056558871678,
+      "loss": 2.7008,
+      "step": 19458
+    },
+    {
+      "epoch": 0.9922816623547981,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00020632174668104296,
+      "loss": 2.7209,
+      "step": 19459
+    },
+    {
+      "epoch": 0.9923326558109035,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00020623838874056635,
+      "loss": 2.7188,
+      "step": 19460
+    },
+    {
+      "epoch": 0.992383649267009,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00020615558211715704,
+      "loss": 2.69,
+      "step": 19461
+    },
+    {
+      "epoch": 0.9924346427231143,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002060733268618946,
+      "loss": 2.7121,
+      "step": 19462
+    },
+    {
+      "epoch": 0.9924856361792197,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00020599162302551788,
+      "loss": 2.7164,
+      "step": 19463
+    },
+    {
+      "epoch": 0.9925366296353251,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00020591047065842612,
+      "loss": 2.7198,
+      "step": 19464
+    },
+    {
+      "epoch": 0.9925876230914304,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0002058298698106781,
+      "loss": 2.7029,
+      "step": 19465
+    },
+    {
+      "epoch": 0.9926386165475358,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00020574982053199245,
+      "loss": 2.7049,
+      "step": 19466
+    },
+    {
+      "epoch": 0.9926896100036412,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00020567032287174765,
+      "loss": 2.7076,
+      "step": 19467
+    },
+    {
+      "epoch": 0.9927406034597467,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00020559137687898188,
+      "loss": 2.7071,
+      "step": 19468
+    },
+    {
+      "epoch": 0.9927915969158521,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002055129826023929,
+      "loss": 2.6933,
+      "step": 19469
+    },
+    {
+      "epoch": 0.9928425903719574,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00020543514009033838,
+      "loss": 2.6955,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9928935838280628,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0002053578493908352,
+      "loss": 2.7321,
+      "step": 19471
+    },
+    {
+      "epoch": 0.9929445772841682,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00020528111055156045,
+      "loss": 2.7193,
+      "step": 19472
+    },
+    {
+      "epoch": 0.9929955707402736,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020520492361985042,
+      "loss": 2.6905,
+      "step": 19473
+    },
+    {
+      "epoch": 0.9930465641963789,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00020512928864270096,
+      "loss": 2.6814,
+      "step": 19474
+    },
+    {
+      "epoch": 0.9930975576524844,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020505420566676767,
+      "loss": 2.7267,
+      "step": 19475
+    },
+    {
+      "epoch": 0.9931485511085898,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00020497967473836535,
+      "loss": 2.7293,
+      "step": 19476
+    },
+    {
+      "epoch": 0.9931995445646952,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00020490569590346852,
+      "loss": 2.704,
+      "step": 19477
+    },
+    {
+      "epoch": 0.9932505380208005,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00020483226920771093,
+      "loss": 2.6954,
+      "step": 19478
+    },
+    {
+      "epoch": 0.9933015314769059,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00020475939469638587,
+      "loss": 2.7306,
+      "step": 19479
+    },
+    {
+      "epoch": 0.9933525249330113,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00020468707241444618,
+      "loss": 2.7216,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9934035183891167,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00020461530240650373,
+      "loss": 2.6832,
+      "step": 19481
+    },
+    {
+      "epoch": 0.9934545118452222,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00020454408471682988,
+      "loss": 2.71,
+      "step": 19482
+    },
+    {
+      "epoch": 0.9935055053013275,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002044734193893553,
+      "loss": 2.7263,
+      "step": 19483
+    },
+    {
+      "epoch": 0.9935564987574329,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002044033064676699,
+      "loss": 2.7277,
+      "step": 19484
+    },
+    {
+      "epoch": 0.9936074922135383,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00020433374599502287,
+      "loss": 2.7303,
+      "step": 19485
+    },
+    {
+      "epoch": 0.9936584856696437,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00020426473801432251,
+      "loss": 2.6789,
+      "step": 19486
+    },
+    {
+      "epoch": 0.993709479125749,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0002041962825681366,
+      "loss": 2.7342,
+      "step": 19487
+    },
+    {
+      "epoch": 0.9937604725818544,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00020412837969869166,
+      "loss": 2.7515,
+      "step": 19488
+    },
+    {
+      "epoch": 0.9938114660379599,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002040610294478739,
+      "loss": 2.7183,
+      "step": 19489
+    },
+    {
+      "epoch": 0.9938624594940653,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000203994231857228,
+      "loss": 2.7059,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9939134529501706,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00020392798696795833,
+      "loss": 2.71,
+      "step": 19491
+    },
+    {
+      "epoch": 0.993964446406276,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020386229482092792,
+      "loss": 2.7049,
+      "step": 19492
+    },
+    {
+      "epoch": 0.9940154398623814,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00020379715545665905,
+      "loss": 2.7098,
+      "step": 19493
+    },
+    {
+      "epoch": 0.9940664333184868,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00020373256891533293,
+      "loss": 2.7044,
+      "step": 19494
+    },
+    {
+      "epoch": 0.9941174267745921,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00020366853523678973,
+      "loss": 2.6941,
+      "step": 19495
+    },
+    {
+      "epoch": 0.9941684202306976,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00020360505446052874,
+      "loss": 2.7148,
+      "step": 19496
+    },
+    {
+      "epoch": 0.994219413686803,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.00020354212662570815,
+      "loss": 2.705,
+      "step": 19497
+    },
+    {
+      "epoch": 0.9942704071429084,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00020347975177114475,
+      "loss": 2.7124,
+      "step": 19498
+    },
+    {
+      "epoch": 0.9943214005990137,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00020341792993531477,
+      "loss": 2.6908,
+      "step": 19499
+    },
+    {
+      "epoch": 0.9943723940551191,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00020335666115635286,
+      "loss": 2.7232,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9944233875112245,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00020329594547205277,
+      "loss": 2.7158,
+      "step": 19501
+    },
+    {
+      "epoch": 0.9944743809673299,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00020323578291986686,
+      "loss": 2.7066,
+      "step": 19502
+    },
+    {
+      "epoch": 0.9945253744234354,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00020317617353690648,
+      "loss": 2.6973,
+      "step": 19503
+    },
+    {
+      "epoch": 0.9945763678795407,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0002031171173599417,
+      "loss": 2.6959,
+      "step": 19504
+    },
+    {
+      "epoch": 0.9946273613356461,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00020305861442540153,
+      "loss": 2.7076,
+      "step": 19505
+    },
+    {
+      "epoch": 0.9946783547917515,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00020300066476937312,
+      "loss": 2.7096,
+      "step": 19506
+    },
+    {
+      "epoch": 0.9947293482478569,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00020294326842760306,
+      "loss": 2.7253,
+      "step": 19507
+    },
+    {
+      "epoch": 0.9947803417039622,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00020288642543549607,
+      "loss": 2.6996,
+      "step": 19508
+    },
+    {
+      "epoch": 0.9948313351600676,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020283013582811593,
+      "loss": 2.7179,
+      "step": 19509
+    },
+    {
+      "epoch": 0.9948823286161731,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00020277439964018484,
+      "loss": 2.7061,
+      "step": 19510
+    },
+    {
+      "epoch": 0.9949333220722785,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020271921690608364,
+      "loss": 2.7008,
+      "step": 19511
+    },
+    {
+      "epoch": 0.9949843155283838,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.00020266458765985196,
+      "loss": 2.713,
+      "step": 19512
+    },
+    {
+      "epoch": 0.9950353089844892,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00020261051193518767,
+      "loss": 2.6775,
+      "step": 19513
+    },
+    {
+      "epoch": 0.9950863024405946,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00020255698976544747,
+      "loss": 2.7231,
+      "step": 19514
+    },
+    {
+      "epoch": 0.9951372958967,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00020250402118364658,
+      "loss": 2.7184,
+      "step": 19515
+    },
+    {
+      "epoch": 0.9951882893528053,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020245160622245863,
+      "loss": 2.6805,
+      "step": 19516
+    },
+    {
+      "epoch": 0.9952392828089108,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00020239974491421576,
+      "loss": 2.6876,
+      "step": 19517
+    },
+    {
+      "epoch": 0.9952902762650162,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00020234843729090868,
+      "loss": 2.7059,
+      "step": 19518
+    },
+    {
+      "epoch": 0.9953412697211216,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002022976833841866,
+      "loss": 2.6884,
+      "step": 19519
+    },
+    {
+      "epoch": 0.995392263177227,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00020224748322535696,
+      "loss": 2.7224,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9954432566333323,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002021978368453857,
+      "loss": 2.7204,
+      "step": 19521
+    },
+    {
+      "epoch": 0.9954942500894377,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020214874427489733,
+      "loss": 2.7078,
+      "step": 19522
+    },
+    {
+      "epoch": 0.9955452435455431,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00020210020554417457,
+      "loss": 2.7087,
+      "step": 19523
+    },
+    {
+      "epoch": 0.9955962370016486,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00020205222068315842,
+      "loss": 2.7234,
+      "step": 19524
+    },
+    {
+      "epoch": 0.9956472304577539,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0002020047897214486,
+      "loss": 2.7073,
+      "step": 19525
+    },
+    {
+      "epoch": 0.9956982239138593,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020195791268830268,
+      "loss": 2.7102,
+      "step": 19526
+    },
+    {
+      "epoch": 0.9957492173699647,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020191158961263694,
+      "loss": 2.7266,
+      "step": 19527
+    },
+    {
+      "epoch": 0.9958002108260701,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020186582052302566,
+      "loss": 2.7056,
+      "step": 19528
+    },
+    {
+      "epoch": 0.9958512042821754,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00020182060544770162,
+      "loss": 2.7125,
+      "step": 19529
+    },
+    {
+      "epoch": 0.9959021977382808,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0002017759444145556,
+      "loss": 2.6925,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9959531911943863,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00020173183745113692,
+      "loss": 2.7176,
+      "step": 19531
+    },
+    {
+      "epoch": 0.9960041846504917,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00020168828458465293,
+      "loss": 2.7167,
+      "step": 19532
+    },
+    {
+      "epoch": 0.996055178106597,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0002016452858419691,
+      "loss": 2.7292,
+      "step": 19533
+    },
+    {
+      "epoch": 0.9961061715627024,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002016028412496094,
+      "loss": 2.7107,
+      "step": 19534
+    },
+    {
+      "epoch": 0.9961571650188078,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00020156095083375577,
+      "loss": 2.692,
+      "step": 19535
+    },
+    {
+      "epoch": 0.9962081584749132,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002015196146202482,
+      "loss": 2.698,
+      "step": 19536
+    },
+    {
+      "epoch": 0.9962591519310185,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0002014788326345849,
+      "loss": 2.7451,
+      "step": 19537
+    },
+    {
+      "epoch": 0.996310145387124,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00020143860490192252,
+      "loss": 2.7164,
+      "step": 19538
+    },
+    {
+      "epoch": 0.9963611388432294,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00020139893144707524,
+      "loss": 2.7134,
+      "step": 19539
+    },
+    {
+      "epoch": 0.9964121322993348,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00020135981229451588,
+      "loss": 2.7061,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9964631257554402,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00020132124746837497,
+      "loss": 2.7112,
+      "step": 19541
+    },
+    {
+      "epoch": 0.9965141192115455,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00020128323699244131,
+      "loss": 2.6881,
+      "step": 19542
+    },
+    {
+      "epoch": 0.9965651126676509,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00020124578089016152,
+      "loss": 2.7122,
+      "step": 19543
+    },
+    {
+      "epoch": 0.9966161061237563,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00020120887918464054,
+      "loss": 2.7328,
+      "step": 19544
+    },
+    {
+      "epoch": 0.9966670995798618,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00020117253189864124,
+      "loss": 2.7135,
+      "step": 19545
+    },
+    {
+      "epoch": 0.9967180930359671,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0002011367390545843,
+      "loss": 2.707,
+      "step": 19546
+    },
+    {
+      "epoch": 0.9967690864920725,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00020110150067454867,
+      "loss": 2.7138,
+      "step": 19547
+    },
+    {
+      "epoch": 0.9968200799481779,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00020106681678027104,
+      "loss": 2.698,
+      "step": 19548
+    },
+    {
+      "epoch": 0.9968710734042833,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00020103268739314628,
+      "loss": 2.7151,
+      "step": 19549
+    },
+    {
+      "epoch": 0.9969220668603886,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0002009991125342271,
+      "loss": 2.6979,
+      "step": 19550
+    },
+    {
+      "epoch": 0.996973060316494,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002009660922242241,
+      "loss": 2.712,
+      "step": 19551
+    },
+    {
+      "epoch": 0.9970240537725995,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002009336264835058,
+      "loss": 2.7146,
+      "step": 19552
+    },
+    {
+      "epoch": 0.9970750472287049,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00020090171533209897,
+      "loss": 2.7232,
+      "step": 19553
+    },
+    {
+      "epoch": 0.9971260406848103,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00020087035878968779,
+      "loss": 2.6992,
+      "step": 19554
+    },
+    {
+      "epoch": 0.9971770341409156,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00020083955687561454,
+      "loss": 2.723,
+      "step": 19555
+    },
+    {
+      "epoch": 0.997228027597021,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00020080930960887956,
+      "loss": 2.7033,
+      "step": 19556
+    },
+    {
+      "epoch": 0.9972790210531264,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002007796170081407,
+      "loss": 2.685,
+      "step": 19557
+    },
+    {
+      "epoch": 0.9973300145092318,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00020075047909171392,
+      "loss": 2.7434,
+      "step": 19558
+    },
+    {
+      "epoch": 0.9973810079653372,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00020072189587757298,
+      "loss": 2.7115,
+      "step": 19559
+    },
+    {
+      "epoch": 0.9974320014214426,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00020069386738334942,
+      "loss": 2.7339,
+      "step": 19560
+    },
+    {
+      "epoch": 0.997482994877548,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00020066639362633252,
+      "loss": 2.6919,
+      "step": 19561
+    },
+    {
+      "epoch": 0.9975339883336534,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020063947462346975,
+      "loss": 2.6976,
+      "step": 19562
+    },
+    {
+      "epoch": 0.9975849817897587,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002006131103913658,
+      "loss": 2.7178,
+      "step": 19563
+    },
+    {
+      "epoch": 0.9976359752458641,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00020058730094628355,
+      "loss": 2.701,
+      "step": 19564
+    },
+    {
+      "epoch": 0.9976869687019695,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0002005620463041437,
+      "loss": 2.7013,
+      "step": 19565
+    },
+    {
+      "epoch": 0.997737962158075,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0002005373464805244,
+      "loss": 2.7187,
+      "step": 19566
+    },
+    {
+      "epoch": 0.9977889556141804,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002005132014906618,
+      "loss": 2.7003,
+      "step": 19567
+    },
+    {
+      "epoch": 0.9978399490702857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00020048961134944986,
+      "loss": 2.7013,
+      "step": 19568
+    },
+    {
+      "epoch": 0.9978909425263911,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00020046657607144004,
+      "loss": 2.7154,
+      "step": 19569
+    },
+    {
+      "epoch": 0.9979419359824965,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020044409567084156,
+      "loss": 2.7206,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9979929294386019,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00020042217016152177,
+      "loss": 2.7007,
+      "step": 19571
+    },
+    {
+      "epoch": 0.9980439228947072,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00020040079955700504,
+      "loss": 2.7206,
+      "step": 19572
+    },
+    {
+      "epoch": 0.9980949163508127,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00020037998387047417,
+      "loss": 2.7319,
+      "step": 19573
+    },
+    {
+      "epoch": 0.9981459098069181,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00020035972311476912,
+      "loss": 2.705,
+      "step": 19574
+    },
+    {
+      "epoch": 0.9981969032630235,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00020034001730238787,
+      "loss": 2.7164,
+      "step": 19575
+    },
+    {
+      "epoch": 0.9982478967191288,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00020032086644548579,
+      "loss": 2.6991,
+      "step": 19576
+    },
+    {
+      "epoch": 0.9982988901752342,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002003022705558762,
+      "loss": 2.7367,
+      "step": 19577
+    },
+    {
+      "epoch": 0.9983498836313396,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00020028422964503008,
+      "loss": 2.7007,
+      "step": 19578
+    },
+    {
+      "epoch": 0.998400877087445,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00020026674372407578,
+      "loss": 2.7333,
+      "step": 19579
+    },
+    {
+      "epoch": 0.9984518705435504,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0002002498128037995,
+      "loss": 2.7068,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9985028639996558,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002002334368946452,
+      "loss": 2.7212,
+      "step": 19581
+    },
+    {
+      "epoch": 0.9985538574557612,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00020021761600671437,
+      "loss": 2.6876,
+      "step": 19582
+    },
+    {
+      "epoch": 0.9986048509118666,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.000200202350149766,
+      "loss": 2.72,
+      "step": 19583
+    },
+    {
+      "epoch": 0.998655844367972,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000200187639333217,
+      "loss": 2.6956,
+      "step": 19584
+    },
+    {
+      "epoch": 0.9987068378240773,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002001734835661417,
+      "loss": 2.6965,
+      "step": 19585
+    },
+    {
+      "epoch": 0.9987578312801827,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00020015988285727193,
+      "loss": 2.7196,
+      "step": 19586
+    },
+    {
+      "epoch": 0.9988088247362882,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00020014683721499742,
+      "loss": 2.7205,
+      "step": 19587
+    },
+    {
+      "epoch": 0.9988598181923936,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0002001343466473654,
+      "loss": 2.7248,
+      "step": 19588
+    },
+    {
+      "epoch": 0.9989108116484989,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0002001224111620806,
+      "loss": 2.746,
+      "step": 19589
+    },
+    {
+      "epoch": 0.9989618051046043,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0002001110307665055,
+      "loss": 2.7033,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9990127985607097,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00020010020546766007,
+      "loss": 2.7182,
+      "step": 19591
+    },
+    {
+      "epoch": 0.999063792016815,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00020008993527222187,
+      "loss": 2.7227,
+      "step": 19592
+    },
+    {
+      "epoch": 0.9991147854729204,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00020008022018652604,
+      "loss": 2.7149,
+      "step": 19593
+    },
+    {
+      "epoch": 0.9991657789290259,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0002000710602165655,
+      "loss": 2.7092,
+      "step": 19594
+    },
+    {
+      "epoch": 0.9992167723851313,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00020006245536799045,
+      "loss": 2.7093,
+      "step": 19595
+    },
+    {
+      "epoch": 0.9992677658412367,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00020005440564610876,
+      "loss": 2.7114,
+      "step": 19596
+    },
+    {
+      "epoch": 0.999318759297342,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00020004691105588604,
+      "loss": 2.7054,
+      "step": 19597
+    },
+    {
+      "epoch": 0.9993697527534474,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00020003997160194517,
+      "loss": 2.6953,
+      "step": 19598
+    },
+    {
+      "epoch": 0.9994207462095528,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002000335872885669,
+      "loss": 2.7238,
+      "step": 19599
+    },
+    {
+      "epoch": 0.9994717396656582,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00020002775811968931,
+      "loss": 2.7238,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9995227331217637,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0002000224840989081,
+      "loss": 2.7292,
+      "step": 19601
+    },
+    {
+      "epoch": 0.999573726577869,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00020001776522947663,
+      "loss": 2.7384,
+      "step": 19602
+    },
+    {
+      "epoch": 0.9996247200339744,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00020001360151430567,
+      "loss": 2.7232,
+      "step": 19603
+    },
+    {
+      "epoch": 0.9996757134900798,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00020000999295596362,
+      "loss": 2.6893,
+      "step": 19604
+    },
+    {
+      "epoch": 0.9997267069461852,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00020000693955667652,
+      "loss": 2.72,
+      "step": 19605
+    },
+    {
+      "epoch": 0.9997777004022905,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00020000444131832763,
+      "loss": 2.7199,
+      "step": 19606
+    },
+    {
+      "epoch": 0.9998286938583959,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00020000249824245824,
+      "loss": 2.7281,
+      "step": 19607
+    },
+    {
+      "epoch": 0.9998796873145014,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00020000111033026687,
+      "loss": 2.7004,
+      "step": 19608
+    },
+    {
+      "epoch": 0.9999306807706068,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00020000027758260952,
+      "loss": 2.6974,
+      "step": 19609
+    },
+    {
+      "epoch": 0.9999816742267121,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0002,
+      "loss": 2.7055,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9999816742267121,
+      "step": 19610,
+      "total_flos": 2.5538897047374004e+20,
+      "train_loss": 1.078860640586608,
+      "train_runtime": 264748.4578,
+      "train_samples_per_second": 66.368,
+      "train_steps_per_second": 0.074
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 19610,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.5538897047374004e+20,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}